This page contains discussions that have been archived from Village pump (technical). Please do not edit the contents of this page. If you wish to revive any of these discussions, either start a new thread or use the talk page associated with that topic.
It has been incredibly slow in saving edits to an article. Going from page to page is fine though. Any idea why? Mcdynamite (talk) 01:29, 23 July 2023 (UTC)
Previewing changes & saving has been very slow since about 12:30 UTC. I thought that if I did something else for an hour it might have come right, but no. Nurg (talk) 01:50, 23 July 2023 (UTC)
It's not just you, and it's not just edits. Pages have been slow to load. Twitter concurs that there's a problem on Wikipedia. schetm (talk) 01:53, 23 July 2023 (UTC)
Thanks for opening the ticket. Same here as the others are experiencing. It's all the wikis - Commons, Wikisource, etc. etc. It was saving OK for me a couple of hours ago, but now it takes forever to save. — Maile (talk) 02:16, 23 July 2023 (UTC)
I recently used Google Chrome to access Wikipedia and the result was "503 Service Unavailable No server is available to handle this request."
Maybe the tool I used for bypassing censorship has an effect, but DownDetector reported that many people are experiencing the same issue. However it's fine when I used Tor browser and the Google Chrome version had just restored. IntegerSequences (talk | contribs) 02:19, 23 July 2023 (UTC)
In the past week or two, but especially in the past couple of days, pages including lots of <math> tags on them have been experiencing excessive latency (10s of seconds) for loading and making edits, and at some times of the day such requests time out entirely (after 60s the server gives up, returning: The maximum request time of 60 seconds was exceeded. [«ID number»] «date string»: Fatal exception of type "Wikimedia\RequestTimeout\RequestTimeoutException"). There has been discussion about this on talk pages for golden ratio, prime number, and possibly other pages, and a slightly more general discussion at Wikipedia:Help desk#Golden ratio. During lower-traffic times of day everything works fine, and also pages seem to load okay for some readers from an Android app. Anyhow, the timeouts are making it very difficult to read / edit articles about technical topics. Some kind of fix would be appreciated! –jacobolus(t)01:33, 22 July 2023 (UTC)
I just viewed Golden ratio (pretty fast), clicked edit (fast), then preview (roughly 80 seconds by wall clock). The HTML source of the preview contained:
I just now tried to visit golden ratio and prime number, and both timed out (60s). When getting a preview of a golden ratio edit, it also took ~80 seconds for me. I tried replacing {{reflist}} with <references> ... </references> and getting another preview, and that didn't speed anything up. Now the relevant section of the source has:
So perhaps reflist is a red herring. (Or maybe there's an issue with rendering the references whether or not it's wrapped in a reflist template.) –jacobolus(t)04:38, 22 July 2023 (UTC)
Hm. It does seems to have something to do with the references. If I remove all of the instances of <ref>...</ref> and {{sfn}} from the article it's much faster to preview. –jacobolus(t)05:04, 22 July 2023 (UTC)
This is probably phab:T341666. I hadn't gotten around to testing if we were experiencing the issue as well, but this section indicates we are. Izno (talk) 05:43, 22 July 2023 (UTC)
Well, the undo now appears in the page history, but not in the article. That is, the most recent revision in the page history is not the current page. Purging the page cache does not fix this. What is going on here? XOR'easter (talk) 02:56, 23 July 2023 (UTC)
Waikato_(rangatira) took only 0.113 seconds in real time while Hooley's delta function took 4.557 seconds, almost 40 times longer. However, the 4103-byte article is less than 6 times as long as the 761-byte Waikato_(rangatira). This is definitely out of proportion.
Please look at WP:RD/Math as it appears currently (i.e. this) when using the default skin in a wide window so that the table of contents is a sidebar. Scroll to the bottom and you will see that the last section heading reads The meaning of . But look over at the sidebar and the table of contents shows The meaning of '"`UNIQ--postMath-00000049-QINU`"' . It ought to be possible to improve on that! --142.112.221.64 (talk) 06:01, 22 July 2023 (UTC)
Or just search for "UNIQ--postMath": [1] (127 hits). It would help if a discussion somewhere decided how to handle these with a link to the discussion in edit summaries fixing the problems. Johnuniq (talk) 09:56, 22 July 2023 (UTC)
I pointed out (fairly facetiously) that the article Victorian Railways Dd class has math markup, including using it in a section heading. I have just looked up that article in the app, and found that the infobox named "Victorian Railways class" (which displays acceptably in a browser), generates a hidden box with the visible title "Quick Facts D E D {\displaystyle \mathrm {D_{{E}^{D}} }" (which, when opened, includes the correctly-rendered name). The use of math markup in the name= parameter is not covered by MOS:HEAD or mentioned in Template:Infobox. I would guess that this is a related issue. -- Verbarson talkedits16:55, 23 July 2023 (UTC)
CentralNotice won't stay dismissed
Over the last three or four weeks, I have been presented with several different WP:CENTRALNOTICE banners. I dismiss them using the "X-in-circle" icon at the top right of the box, yet they keep on returning, sometimes several times in one day. For example, the banner reading "Join the Wikipedia Pages Wanting Photos Campaign 2023 to help improve Wikipedia articles with photos and win a prize!" (the id for which is wikipedia_wanting_photos_2023) has been shown to me three times today - twice in my session of 08:00-08:45 (UTC), and once in the session that I began at 16:45 (UTC). I can hide them all permanently by unsetting everything at Preferences → Banners, or by setting "Suppress display of all CentralNotices (To suppress only certain classes of notices use the Banners option in preferences)" at Preferences → Gadgets, but these methods then won't allow any banners to be displayed, not even new ones that I might be interested in. Short of doing this, which non-tech-savvy readers may not be happy doing, how can banners that are no longer of interest be dismissed permanently and yet still allow a completely-new banner to be displayed? --Redrose64 🌹 (talk) 17:34, 23 July 2023 (UTC)
Image rotates 90 degrees automatically with flipped aspect ratio
This is the image.
For whatever reasons the image on the right doesn't appear like it actually is. Click it to open in media viewer and it is okay. On the page, it is rotated 90 degrees anti-clockwise. Not just that, the height-to-width ratio is completely flipped so that the image now looks very squeezed. I saw this happen at Songkhla but the issue appears on every page, including here and at File:Naga Head.jpg#filehistory. But scrolling to the top of file page, and it is perfectly fine. What is happening here? —CX Zoom[he/him](let's talk • {C•X})18:04, 23 July 2023 (UTC)
It was, I think the technical term is, screwed up. Some parts of the system thought it was rotated, and some didn't. I re-uploaded it, and after a bunch of purging, it seems to be displaying correctly. – Jonesey95 (talk) 18:25, 23 July 2023 (UTC)
This generally happens if different metadata systems (images can have multiple) have it marked as differing rotation. The information is then conflicting and results become unpredictable. —TheDJ (talk • contribs) 19:11, 23 July 2023 (UTC)
More articles with misplaced infobox on mobile version
Which shouldn't cause an issue per the discussion in the archive, now that I've readded the relevant class to the container. Izno (talk) 19:14, 23 July 2023 (UTC)
Makes sense now. But, considering that campaignbox is not visible on mobile anyway, is it possible to selectively apply stack only on desktop mode? Maybe by using a variant of stack template that uses the same class as infoboxes, so mobile view remains consistent? —CX Zoom[he/him](let's talk • {C•X})19:20, 23 July 2023 (UTC)
If the JavaScript of the form can check the user's groups or privileges, it could simply drop back to the default editor whenever a lack of skipcaptcha is detected. ~ ToBeFree (talk) 00:06, 24 July 2023 (UTC)
The right fix here is to detect if the edit API response is asking for a captcha (if so, it also provides the URL to the captcha image), show the image on the UI and ask the user to solve it; and then retry the edit API with the captchaid and captchaword fields. We already handle this in the WP:SUBMIT script (see MediaWiki:AFC-submit-wizard.js). Someone should implement it here as well – much of the logic can be copy-pasted. – SD0001 (talk) 08:04, 24 July 2023 (UTC)
Not digging the new mobile diffs
The most recent Thursday update has changed the highlighting on mobile diffs such that instead of highlighting the whole block of text including spacing between lines, now the highlighting only goes as high as the text, leaving dark grey bars in between each line.
For some reason, this makes it way harder for my brain to read. I find myself taking more than twice as long to read a block of text, and have to reread lines frequently, almost like I'm reading it upside down. I never knew I was so dumb.
Is there any CSS I can add somewhere to revert to the old highlighting? Or alternatively, to force standard diff view on mobile? I have to swap into it about a third of the time anyway, and maybe this highlighting change is the universe's way of telling me to stop using mobile diffs altogether. Folly Mox (talk) 11:46, 21 July 2023 (UTC)
Do you have examples of diffs you find hard to read? If you can successfully illustrate the difficulty, that might help devs change their minds. Nardog (talk) 13:53, 21 July 2023 (UTC)
I mean, I suppose I could link the most recent diff on my watchlist, Special:MobileDiff/1166429149, but my reading problem universal to all examples. It's like the dark grey bar is guiding my eye to the wrong line, and I end up reading the line I just read instead of continuing on to the next one. Folly Mox (talk) 14:00, 21 July 2023 (UTC)
No changes to mobile diffs were intentional, however this change was intentional for the upcoming inline diffs for desktop. We were nervous folks wouldn't like this particular aspect, and I guess this is confirmation. Allow me to explain the logic, however; phab:T327645 is the relevant task. As you can see, the concern was that the background extending across to the other edge could be misinterpreted as added newlines, or at the very least, is indistinguishable from someone who added a bunch of spaces that happened to stretch across. I hope my explanation makes sense.
I wasn't able to find a happy medium where we use inline elements (fixing the aforementioned issue) but also retaining the background color between each line of text. It might be possible, but assuming it's not, which do you think is better? cc @Folly Mox but inviting input from everyone.
I should note that there will soon hopefully be less confusion on where newlines are, as we will clearly indicate those (phab:T330247).
User:MusikAnimal (WMF) please don't apologise to me or allow my personal experience to guide development. I'm just one data point, and my brain is super weird.[original research?] It may very well be the case that this is a total non-issue or even improvement for literally everyone else. I notice there are (carriage return symbol) newline elements in the inline diffs now, which seems like it might help solve the issue. I have unrelated suggestions I'll drop off at the meta page in a moment. Folly Mox (talk) 22:13, 21 July 2023 (UTC)
Oops just kidding the meta page is just about paragraph breaks, and I don't really know how to interact with phabricator. My suggestion was to remove - and = from the list of characters that are interpreted as word breaks, so inline diffs showing changes in templates look less garbled. Folly Mox (talk) 22:22, 21 July 2023 (UTC)
I'm not sure if I'm seeing the same, but I struggle reading the new diffs too to the point where I cannot finish reading a relatively small diff like Special:MobileDiff/1166705972. I don't see dark grey lines between the green highlighting, but white lines. As such, the contrast between green and text is less than the contrast between white and green. I think this can only work with quite a light green colour. User:MusikAnimal (WMF), thanks for considering rolling it back. —Femke 🐦 (talk) 07:49, 23 July 2023 (UTC)
Hello. I posted my question in Teahouse and they redirected me here since this place has many professionals regarding technical questions. I have question regarding visual editing. I specialize in filmography editing and I noticed a slight problem. Whenever I am adding info through source editing and I want to look how it looks through visual editing, It wont let me. I click visual editing button when I am in source editing domain and it doesnt work. I dont know whether this is bug or new update, but after Thursday, it doesnt work. I can look through "preview" but its annoying since I cant fix anything right on spot. So if you have any tips or info, I will be glad. Ps, Im editing on Android and mostly on Google Chrome. Kesseder (talk) 16:56, 24 July 2023 (UTC)
Do you have an example of a section where this has happened? What exactly happens when you click "Visual editing"? (Any error message, does the progress bar show up, etc.) Is "Use the wikitext mode inside the visual editor, instead of a different wikitext editor" in Preferences on? Nardog (talk) 22:05, 24 July 2023 (UTC)
Basically, nothing happens when i click visual editing. Source editing works normally. For example, I want to change something in source code and I want to see how it looks (through visual editing), it wont let me, the button is not working for me. I click it and nothing happens (no error, no message). I tried it even on my editor through my user page, I write something on source editing domain, i cant see how it looks visually, so I dont know whether this is something new (like update) or its bug, but its confusing and annoying for me. Thx in advance for any additional info Kesseder (talk) 22:49, 24 July 2023 (UTC)
It's indeed a bug introduced last week, and affecting the mobile editor only: T342490. Sorry about that; hopefully it will be fixed soon. In the meantime, if you can bear it, try editing using the desktop editor (you can access it on mobile devices by the "Desktop" link in the page footer, and switch back using the "Mobile view" link in the footer). Matma Rextalk22:50, 24 July 2023 (UTC)
You have run over the transclusion limit. The only fix is to transclude less content. I've used the bare reference tag as a start. Removing the use of {{font}} would be the next place to work on the problem. Izno (talk) 23:04, 23 July 2023 (UTC)
You should also consider not having a 600k article. That would be the other issue. Use some editorial discretion to remove old or insignificant data. Having poll results from 2018 for an election not occurring until this year is somewhere in the realm of unnecessary, at best. Izno (talk) 23:26, 23 July 2023 (UTC)
I appreciate the feedback Izno, and your fix has made things better but there's still an issue. FWIW, I only came across the problem via a Teahouse question. I haven't edited the article other than to add a {{Very long}} template to its top. Before asking about this here, I posted {{Please see}} messages on various Wikipedia Project talk pages and also a few times on the article's talk page, but that's been the extent of my involvement. So, the "you" in this case is not "me", but rather all of the others who've been expanding the article, and continue to expand the article even after your "fix". Would you mind if I copied-and-pasted your above comments onto the article's talk page? Is there a way to WP:PP the article to temporarily stop it from being further expanded until the issue can be resolved through discussion on the talk page? -- Marchjuly (talk) 03:14, 24 July 2023 (UTC)
I see you did a WP:SPINOUT, which seems like a good technical solution. However, these articles may run afoul of WP:NOTSTATS. Articles should be mostly prose and references. Whenever we have a situation where an article is mostly stats, that is a red flag. –Novem Linguae (talk) 08:22, 24 July 2023 (UTC)
@Novem Linguae: The SPINOUT was done by one of the main contributors to the article, not myself. That seems to have resolved the TLIMIT issue. Quite a lot of the {{font}} templates were removed by a different user and that too seem to resolve the problem, but that edit was reverted and the article was then spun off. I don't have much experience in the editing of these articles and perhaps the intention is to eventually add more prose; so, I guess any NOSTATS concerns and anything else probably should be raised on the article's talk page. Finally, thanks to you and everyone else who responded here or on the article's talk page, and helped sort this out. -- Marchjuly (talk) 00:19, 25 July 2023 (UTC)
Tech News: 2023-30
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
The new version of MediaWiki will be on test wikis and MediaWiki.org from 25 July. It will be on non-Wikipedia wikis and some Wikipedias from 26 July. It will be on all wikis from 27 July (calendar).
How can I use a fair use image in multiple related articles at the same time?
The image File:City of Yellowknife CoA.svg is the emblem of the City of Yellowknife and is protected by copyright until 2050 because it is protected by Canada's Crown Copyright Act (meaning the image cannot be exported to Wikimedia Commons until 2050). Currently the image is only used in the article introducing the city of Yellowknife, but another article Coat of arms of Yellowknife introduced the design of the logo in detail, so I need to use this fair use image in an article that is closer to its theme. Fumikas Sagisavas (talk) 04:09, 25 July 2023 (UTC)
Hi Fumikas Sagisavas. For future reference, it's generally better to ask questions of this nature at Wikipedia:Media Copyright Questions since this noticeboard is more for asking technical related questions about things Wikipedia. I'm guessing your question is related to this edit made to Coat of arms of Yellowknife by JJMC89 bot, but please clarify if it's not. That particular bot has been tasked with looking for non-free files being used in a manner that doesn't comply with Wikipedia non-free content use criterion #10c. Each use of a non-free file is required to meet all ten of the non-free content use criteria listed here. One of these criteria (more specifically one part of one of these criteria) is #10c, which states that a separate, specific non-free content use rationale which clearly indicates where the file is being used and how its use satisfies the aforementioned 10 criteria be provided on the file's page for each use of the file. Since there is no rationale for the file's use in the COA article on the file's page, the bot removed the file and left an edit summary linking to WP:NFC#Implementation explaining why. If you feel the file's non-free use in the COA article complies with relevant Wikipedia policy, you should add a non-free use rationale for the file's use in the COA article to the file's page and then re-add the file to the article. This should stop the bot from removing the file. There are, however, a couple of things to consider before you do that. The first one is that adding a non-free use rationale for a particular use doesn't automatically mean the use is policy compliant as explained here; the use needs to clearly meet all 10 criteria for it to be considered valid. This can be tricky when it comes to COAs per WP:FREERbecause the written description of the COA (i.e. the blazon) is generally not considered to be eligible for copyright protection and it's only the graphic representations of that description (i.e. the emblazon) which tend to be protected by copyright. In other words, a free equivalent version of COA might be possible for someone to create by following the instructions given in the blazon. The next thing is that it's not clear why a stand-alone article is needed for that particular COA. At first glance, the COA doesn't seem to meet any of Wikipedia's notability guidelines to justify a stand-alone article. The content about the COA might encyclopedically relevant enough to include as a section in the article about the city, but it might not be enough to warrant its own separate article. So, you might be better off asking about such things at Wikipedia talk:WikiProject Heraldry and vexillology because that's where you're likely going to find Wikipedians interested in and more familiar with articles about COAs. One of the members of that WikiProject might be able to help answer both questions about the file's non-free use and about the Wikipedia notability of the COA. Once the COAs notability has been sorted out, figuring out how to best add the image can be discussed. -- Marchjuly (talk) 05:51, 25 July 2023 (UTC)
Massive wide-ranging IP block on Airtel India users
I get notified when a new incoming link is created to a disambiguation page I created.
Is there a way to do the same for groups of dab pages? I. e., can I watch a list of pages for new incoming links? Paradoctor (talk) 21:02, 25 July 2023 (UTC)
@Certes Not quite what I had in mind, but the Toolforge report will do. Thanks!
Related Changes does not show new incoming links, only changes to the pages themselves.
Is there a reason there is no bot task that reroutes direct links to dab pages without primary topic? Would seem a no-brainer, IYAM. Paradoctor (talk) 21:46, 25 July 2023 (UTC)
Use {{cite av media}} in the Works cited section, and link to it with {{sfn}} using "|loc=time stamp" rather than the "|p=" parameter. Templates like {{sfn}} or {{harv}} aren't valid references on their own, they need to link to a cite containing the full details. That's why you have to have the {{cite av media}} as well. If you have questions about specific examples feel free to ask on my talk page. -- LCU ActivelyDisinterested∆transmissions∆ °co-ords° 18:44, 26 July 2023 (UTC)
Context-significant numbers in popups
If you've activated article-link popups, mouse over a link to Rail transport in Laos and you're informed that "Laos has of standard-gauge railways, primarily..." Visit the article, and you read that "Laos has 422 km (262 mi) of 1,435 mm (4 ft 8+1⁄2 in) standard gauge railways, primarily..."
I understand the importance of removing parenthetical details from popups, but is there a way to include numbers by default? Ideally, the popup would say "Laos has 422 km of 1,435 mm standard gauge railways, primarily..." Presumably, the point of popups is to get a brief preview of an article without visiting it, so when an omission makes the popup look like a mistake ("has of"), something ought to be modified. Nyttend (talk) 21:37, 24 July 2023 (UTC)
A very old bug / feature: templates, or most templates, are excluded from rendering of the pop-up window. See T152639 for some discussion. I think, but I am not sure, that selecting "Enable page previews" from the Appearance preference tab and disabling the pop-ups gadget may work better for you. I don't use either of them, but I just turned on page previews, and the convert template renders fine. – Jonesey95 (talk) 22:20, 24 July 2023 (UTC)
When I mouse over your link above, I see: ""Laos has 422 km (262 mi) of 1,435 mm standard gauge railways, primarily...", i.e, it's doing the first one but not the second one, and given what's been said so far, I'm not sure why I'm seeing that, but I am. I checked the article, and no one has changed it since you wrote. Mathglot (talk) 01:47, 28 July 2023 (UTC)
You're using the now-"standard" popups. There is a gadget which is basically ancient that does things much differently as a result of its age. We're talking about the second one. (See PrimeHunter's link.) Izno (talk) 01:57, 28 July 2023 (UTC)
Can somebody point me at the source repository (and, ideally, the source file) for the parser responsible for processing reply comments on Talk pages? I'd like to take a swing at convincing it to not automatically indent reply lines while inside <syntaxhighlight>...</syntaxhighlight> tags, because man is having to fix that getting really old. FeRDNYC (talk) 02:34, 28 July 2023 (UTC)
Some pages are stuck in this category after move. They are obviously added there by some template, but I can not find which one. Could someone help please. Thanks. Ymblanter (talk) 15:32, 27 July 2023 (UTC)
Have I misremembered it, or do we have a template for our event pages (for meeetups, editathons, etc.) that generates an iCalendar file which users can use to add the event to their calender?
Thank you, all. The template on Meta is what I was thinking of. It is good, but could be developed further such as the inclusion of timezones (now that we have so mny online events). The functionality is sorely eneed on this and other projects. Would anyone be inetrested in working with me to develop it? @WOSlinker and Jdforrester: who were invlved in the original template or its subtemplates. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits13:11, 28 July 2023 (UTC)
Watchlist bug?
My watchlist has started to behave oddly. Entries are no longer collapsed, with a blue triangle next to each title in order to expand it. Instead, there is a blue circle next to each title, and all changes appear with no possibility of collapsing items. This makes viewing the watchlist cumbersome, particularly if there are many changes to any one page.
This behaviour has only started today; previously the page worked as intended. And when I open the watchlist, it very briefly displays correctly, then automatically expands all entries. I have not made any changes to my settings or preferences. RolandR (talk)20:57, 20 July 2023 (UTC)
Mine's behaving strangely on my laptop. I have user:js/watchlist.js, which enables the sorting of watchlist. When I go on my watchlist, it first loads [2], then when the sorting arrows would appear, the box disappears entirely [3].
If I'm on my desktop, things behave normally (actually, there's a different issue there, every collapsed list of changes is expanded Headbomb {t · c · p · b} 08:12, 21 July 2023 (UTC)). I can't find any differences in configuration on either of my computer, both use the same browser (Firefox, same version). Maybe it's related? Headbomb {t · c · p · b}21:40, 20 July 2023 (UTC)
watchlist displays gray collapse arrows even when new, unvisited changes are present in the collapsed list; this despite "Display green collapsible arrows and green bullets for changed pages in your watchlist, page history and recent changes (unavailable with the improved Watchlist user interface)" (quoted from gadgets:Watchlist; a system default; and I am not using the 'improved' watchlist interface).
This smells like a someone deciding to fix something that was not broken.
It's started earlier in the week for me. I want to say Monday ish, but it could have been +/- a few days before/after. Headbomb {t · c · p · b}21:53, 20 July 2023 (UTC)
When not in safe mode (https://en.wikipedia.org/wiki/Special:Watchlist), at the top of the watchlist this: "Pages that have been changed since you last visited them are shown with a green marker." This is partially true; the expand/collapse arrow markers are always gray; never green as they should be when there are unvisited changes hidden inside the collapsed watchlist entry. Earlier on Thursday, the expand/collapse arrow markers were green when unvisited changes were hidden inside the collapsed watchlist entry. The markers for unvisited single-watchlist entries (only one edit so no need for a collapse) are green bullets; after visiting a green-marker-change, the marger goes gray (this is as it should be and has been for as long as I can remember).
When in safe mode (https://en.wikipedia.org/wiki/Special:Watchlist?safemode=1), at the top of the watchlist this: "Pages that have been changed since you last visited them are shown in bold with a green marker." Pages that have changed since my last visit are shown in bold. There are no the green bullet markers for unvisited single-watchlist-entries and no gray bullet markers for visited single-watchlist-entries. The expand/collapse arrow markers are always gray regardless of the bold (unvisited) or not-bold (visited) changes in the watch list. So, yes, a variant of the issue happens in safe mode.
are shown with a green marker is a customization specific to English Wikipedia and should not be expected in safe mode. Izno (talk) 04:37, 21 July 2023 (UTC)
I'm seeing this problem in Timeless but not in any other skin. I have the following preferences enabled: expand watchlist to show all changes, use non-Javascript interface, group changes by page. I've tried unsetting and resetting these as suggested above, but it makes no difference; also makes no difference whether Javascript is on or off.Yesterday's WP:THURSDAY update includes the following:
Yes, that's probably it. It removes a few class names and changes some element types that were likely being used for some of the above customizations. Izno (talk) 16:11, 21 July 2023 (UTC)
For me, the unwanted behaviour only occurs when refreshing; clicking the Watchlist link or navigating via a browser bookmark works normally. If you usually keep your watchlist open and refresh it occasionally, it's worth trying another way. Certes (talk) 17:50, 28 July 2023 (UTC)
What happens if there is an error in the title of a ref?
I intended to correct a typo that I found when I made a typo doing a search. But when I checked the ref, the typo was in the title. So I tried to add "sic" so it would be clear that this didn't need to be corrected. Didn't work. — Vchimpanzee • talk • contributions • 22:36, 19 July 2023 (UTC)
Not a WP:VPT topic. Most typos like that should simply be silently corrected. See MOS:TYPOFIX.
It has been suggested that these be simply corrected however many editors don't like such corrections being made. The best argument I have seen not to make corrections is that someone may wish to use the original text as a search string. Consequently if there have been corrections to faulty spelling then this makes such searching problematic. What I do on ref titles is use an alias of 'notatypo' which is 'as written'. Personally I feel that when a misspelled word has this or say a 'sic' template applied then the word should be obfuscated, always. So, I would typically apply the template as follows {{as written|rec|ieve [sic]}}. This entry format does not cause issues with the CS1/2 template formats which is what the 'sic' template does. You will see here a warning requesting that the sic template not be used in CS1/2 formats (.....because it includes markup that will pollute the COinS metadata.). Regards, Neils51 (talk) 10:18, 20 July 2023 (UTC)
If the typo is in the original, then the typo should be reproduced. You can add a comment like Ice craem<!-- typo in original --> is toxic, says researcher. Headbomb {t · c · p · b}21:46, 20 July 2023 (UTC)
You can use {{sic|nolink=y|reason=error in source|speling eror}} inside ref templates or references. Please don't obfuscate the spelling errors, as it causes extra time wasting, that would be avoided with a [sic] attached. Another thing to watch out is whether the digitised or index web page has the error, but the original does not. If a scan of the original is available then you can use that correct title, if it is correct. Graeme Bartlett (talk) 22:52, 20 July 2023 (UTC)
Yes, 'sic' with 'nolink=y' will work however some editors will point out that within the sic template documentation there is a warning not to use 'sic' in CS1/CS2 citation templates. Good luck with arguing that one. Time wasting? yes, when tools pickup non-obfuscated misspellings. Neils51 (talk) 13:08, 26 July 2023 (UTC)
I think it's really bad to fix typos in reference titles: the whole point of specifying a ref title instead of just a bare URL is that people can use it to find the ref in the event that the URL is inaccessible. jp×g17:20, 29 July 2023 (UTC)
Text size issues on Edge
Sorry to be a double poster. Randomly today the text size of articles here and some other sites (but not many) has randomy changed, making the body text small and the references bigger. It seems to be primarily body text and references because it's not affecting media captions, bullet pointed cast lists, infoboxes, TOC, or External Links. I've tried resetting Edge to default and I've tried futzing with the options under Appearance but i can't make the text normal while simultaneously shrinking the references back down and because the refs are larger it creates extra line space between text. I don't have any custom CSS apart from additions to display error messages, does anyone have any idea how I can fix it? It's absolutely Edge because it looks normal under Firefox. Darkwarriorblake (talk) 21:15, 26 July 2023 (UTC)
Is the zoom level normal? Does changing it help? Which version of Edge are you running? Does logging out fix it? (if you want to test vector2010 while you're logged out click here) --Chris08:46, 27 July 2023 (UTC)
The zoom level is 100%, I tried logging out but it did the same thing. What I have noticed on this discussion is that my initial text is small, but the subsequent responses, being indented, are normal size. It's bizarre, like it has to be some form of CSS change, I can't think what else would specifically affect only certain text, but it's the same across other websites. Darkwarriorblake (talk) 09:56, 27 July 2023 (UTC)
Nope not that either, removed everything, I only wish I could uninstall Edge and start from scratch. I don't want the hassle of switching browsers but it's bizarre why it's doing this. Darkwarriorblake (talk) 21:07, 27 July 2023 (UTC)
Also for further info, if I refresh the page I can see for an instant that the text is normal before it's switched to small, it's some kind of setting actively changing it. Darkwarriorblake (talk) 21:10, 27 July 2023 (UTC)
@Darkwarriorblake: You don't have any custom CSS, but is any weird CSS being applied to the content? Hit F12 when you're on a page with weirdly-sized content and see what CSS rules the inspector is showing for that text, then chase those rules back to their source (wherever it may be). FeRDNYC (talk) 01:27, 28 July 2023 (UTC)
Brill idea FeRDNYC, I looked (I'm not an expert), and this is what I get. The first is text in the bulleted lists which to me appears normal, the second is the text in < p > which is the one that appears small.
The font-size: 12px may be the cuplrit. Can you see what file it is getting that styling from (there should be urls on the right hand side)? --Chris09:50, 28 July 2023 (UTC)
I realized I can untick the 12px and it does change the font size in < p > although it's not affecting the oversized references and I can't see any font-size related info attached to those elements. I've uploaded this PDF with some screenshots of what I can see (https://file.io/fNKU2XFgqLyK) but I cannot see anywhere to follow the 12px back to its source? Darkwarriorblake (talk) 10:23, 28 July 2023 (UTC)
Does the system where this is occurring have a HiDPI[a] screen, by any chance, Darkwarriorblake? Typically, on displays with more-than-HD resolution, screen elements get scaled so that any measure expressed in raw numbers of pixels will actually represent a greater number of true, physical pixels. (Because everything would be far too tiny to see, if length values were taken literally.) If Edge is on a scaled display, but its font rendering isn't being scaled for some reason, that might cause a greater-than-usual disparity between font sizes expressed in pixels (like 12px) and ones expressed using other units (like 0.875em), and that disparity could be why you're seeing text rendered at unexpected sizes (in both directions).
But as far as other text, like references, goes... Actually, when you say "references" are larger than normal, what are you referring to? The references list at the end of the article? The small footnote markers[like this] within the body text? Or something else?
...Regardless, every piece of content has at least some minimal CSS associated with it; if nothing else the browser has a baseline set of default styles that will always be applied. (Though it's possible a setting in the inspector is hiding those by default.)
Style rules are technically applied to HTML elements (tags, basically), rather than the text within those tags, so to find the CSS that governs some text you often have to examine whatever is the most immediate enclosing HTML element. (In Wikipedia content that will usually be a pair of <p>...</p>, <li>...</li>, <span>...</span>, or sometimes <div>...</div> tags. Or, inside of a <table>...</table>, the actual text content will be enclosed in, and styled by, the inner <th>...</th> and <td>...</td> tags.) FeRDNYC (talk) 02:28, 29 July 2023 (UTC)
It's only 1920 x 1980, not sure what the max resolution is but the monitor has been fine for a year. It's definitely something in the browser but I can't get rid of the bloody thing to start afresh. I think I will just have to switch browsers. In terms of references I mean the in-line ones, the numbers/letters. Darkwarriorblake (talk) 19:49, 29 July 2023 (UTC)
^HiDPI is any display with a native resolution greater than 1080p (full HD), like a 2.5k, 4k, 5k, or 8k screen. Just to be confusing, the display industry decided to switch from vertical to horizontal pixel dimensions when discussing screen resolution, post-HD. So, 1080p == 1920px wide, or "2k" in the newer parlance. You can usually find your screen dimensions and any scaling configuration in the Display section of the Settings app.
WikiProject banner is broken and I can't figure out why/how
The project's custom importance mask was emitting an extra space which was confusing the template. I've fixed it now — Martin (MSGJ · talk) 10:21, 30 July 2023 (UTC)
Can anyone tell what this fat: project is for and when it was created?
How to recover the two English wiki pages, one of which is a redirect and the other appears to be an article?
I know I should probably make a Phabricator about this, but starting here since that's easier and less bureaucratic. wbm1058 (talk) 12:22, 30 July 2023 (UTC)
The article looked like this on 30 November 2016 and was never scraped again by archive.org, so it may have been deleted (I wouldn't be surprised, looking at the state of it). Black Kite (talk)12:57, 30 July 2023 (UTC)
I was just about to create a new thread about this. The aforementioned page moves did not seem to include the talk pages, so [[Talk:Fat: An Appreciation of a Misunderstood Ingredient, with_Recipes]] and [[Talk:Fat: The Fight of My Life]] are also facing the same issue. Pinging Taavi. ✗plicit13:51, 30 July 2023 (UTC)
This made me curious - as of the most recent database dump, there are ~5300 mainspace pages on enwiki that match the pattern XYZ: Something or XYZ:Something, with ~650 different three-letter groups. Of those, 2059 are mos:, 558 cat:, 457 std:, 241 atp:, and 128 csi: - the rest are used under 50 times and more than half are only used once.
There is an incubator wiki for Mooré (mos:) and it is plausible it will be assigned at some point in the future - it seems to be reasonably active. Might be worth thinking now about how we plan to address this, since I think our use of things like MOS:NUMBERS is pretty prolific.
One additional check: fat: seems to be the only prefix on Special:Interwiki that has colliding enwiki page titles, and it was only the two that have been fixed here, so we don't have any others out there waiting to trip us up. For the moment... Andrew Gray (talk) 14:20, 30 July 2023 (UTC)
Haven't heard anything about this for a while, but as far as I can tell, the little oopsie-woozie from a few months ago seems to still be in full force (graphs are not rendering properly anywhere). Does anyone know what the deal is on this, or what progress is being made? Or what issues need to be resolved to move forward? jp×g17:18, 29 July 2023 (UTC)
@JPxG: Graphs are not broken, they are deliberately disabled because a security issue was found (there are several threads in Archive 205 and Archive 206 on the matter). At the time that they were disabled, thry were working as normsl. There is more at phab:T334940, to which you can comment or subscribe. --Redrose64 🌹 (talk) 10:43, 30 July 2023 (UTC)
What I mean by "busted" is that previously a software package could be used for its intended purpose, and now it cannot. Similarly, I would call a drill press "busted" if it had a wiring fault that electrocuted the operator when turned on, regardless of whether the motor was still able to rotate the spindle. jp×g20:49, 30 July 2023 (UTC)
To provide an alternative, at Category:Orphaned articles, I added plain horizontal bar graphs for tracking to help with WP Orphan. You're welcome to use something like this elsewhere to visually show progress. Regards, JoeNMLC (talk) 21:12, 30 July 2023 (UTC)
different placing within a table's single cell
In this collapsed table, the currently ongoing World Championship in Women's football is at bottom. Because it is held in 2 countries 2 flags are shown. The 2 flags are placed horizontally thus widening the column. I would like to place them vertically to reduce the column width, but I don't want the text "2023 WC" to follow either of those flags. Is there any way I can get "2023 WC" to stay vertically aligned centrally next to the flags? Dutchy45 (talk) 05:52, 31 July 2023 (UTC)
@Dutchy45: Columns might work but maybe read oddly in screen readers, and be a bit odd in general. I used style="position: relative; top: -0.7em;".[4] It's centered in my browser but may not be exact in all circumstances. PrimeHunter (talk) 09:31, 31 July 2023 (UTC)
Please delete the overrides for these three localisation messages. They were created a while ago as a workaround for the bug T301203, which has since been fixed, but global interface editors such as myself and Jon aren't able to delete pages. They result in incorrect formatting in some error messages: for example, when you can't move a page (try viewing Special:MovePage/Test while logged out), it says "You do not have permission to Move" instead of "You do not have permission to move this page". These three are the only inconsistent messages (see [5]). Matma Rextalk10:50, 31 July 2023 (UTC)
how to determine the current protection level of a page
What's the quickest way to determine the current protection level of a page? An admin can click "change protection" and the resulting page shows the current protection level (at least on Wiktionary), but what is a non-admin to do? The best method I've found so far is the "View logs for this page" link on the "View history" tab, but for e.g. Fascism, determining the current level requires working out whether the latest entry in that log reflects the current state, which in this case it doesn't: the latest entry there was removing Pending Changes protection, but the page is not unprotected: if I look further down the history, past some visibility changes that don't affect the protection level, I see where years earlier in the log, autoconfirmed protection was applied and has not expired because it was indefinite — on other pages, where protection was for a definite period, I have to compare the expiry date to the current date and UTC time to check if it's still in effect. And for entries like Gamergate (harassment campaign), the log just says the protection was moved, without even saying what it was or is. Is there a page or magic word that just returns something like "the following protection level(s) currently apply to page X: 'autoconfirmed', 'pending changes'"? -sche (talk) 22:55, 29 July 2023 (UTC)
Oh, that's very useful, thank you! I had considered suggesting that in the same way a user has the option to "view source" even when the normal "edit" button is gone due to page protection, it might be useful to have a "view protection" button even when the user is not an admin and so cannot "change protection"; it's great to know that such a functionality already exists! Since there wasn't a button, I didn't think to try navigating to it manually. -sche (talk) 00:31, 31 July 2023 (UTC)
This method has another advantage, which is that it also shows the Pending Changes (PC) level which might not appear elsewhere. Compare Special:Protect/1984 with Special:PageInfo/1984 (including the logs). I think it should head for the type of functionality you see at Special:UserRights, which differs far more depending on whether you're an admin. We link that in the side bar. For some wikis it may be appropriate (and easy) to link protection there, but I don't think it's worthwhile here. Still, it's a handy link to know. -- zzuuzz(talk)12:51, 31 July 2023 (UTC)
MS Word to wikitext?
I was helping a new editor yesterday at an edit-a-thon. It was really an eye-opening experience as to just how steep the learning curve is.
The immediate problem was that she had written a first draft in MS Word, then copy-pasted it into her sandbox. Not surprisingly, it made a total mess of the formatting. This seems like a common enough scenario that we really should be able to do better. Is there a tool to take a MS Word file and import it, converting the formatting into some reasonable wiki markup? RoySmith(talk)15:18, 30 July 2023 (UTC)
VE does a good job at it. I made a quick test to make sure: User:Nux/test_office.
One problem is doing proper formatting in Word or other office software. In my experience as an IT guy, with quite a lot of documents passing through my hands back in the days, people don't do formatting. They do bold instead of proper headers. They do all kinds of marks instead of proper enumeration and they do double enter instead of proper paragraphs. With proper, structural/semantic formatting in office you should mostly be fine. Just use VE and paste. Nux (talk) 15:50, 30 July 2023 (UTC)
Increase paragraph spacing to be greater than line spacing, so that you don't need to type an empty line between paragraphs. I don't have Word installed right now, but the last time I used it, this was actually how it behaved by default. In Google Docs, you'd use the toolbar item "Line & paragraph spacing → Add space after paragraph", and then "Normal text → Update 'Normal text' to match". In LibreOffice Writer, you'd right-click some text, then use "Paragraph → Edit Style… → Indents & Spacing → Spacing → Below paragraph". Matma Rextalk10:30, 31 July 2023 (UTC)
Yeah, people often use space for indent and enter to increase spacing. I guess it's easier that way as you don't need to setup paragraphs in windows with lots of options. I know that a few people I asked about it didn't even know that there is a better way to do documents. Nux (talk) 19:53, 31 July 2023 (UTC)
Article previews are broken when the article has coords in it
Screenshot of preview when the article has a coords template in it.
Articles that have coordinate templates, such as Mercedes AMG High Performance Powertrains, have broken preview popups when you hover over the link – there won't appear any text / the text is blank. Sometimes the image will render but the text won't. See the first screenshot.
Screenshot of preview when the article does not have a coords template in it.
Articles that don't have coordinate templates on the other hand, like for example 2023 Australian Grand Prix, will have preview popups render correctly. See the second screenshot.
I am using Google Chrome 114.0.5735.248 official build 64-bit, on Windows 10 Pro 22H2 64-bit 19045.3208. The screenshots were taken in incognito mode (to isolate extensions, scripts etc as a possible cause), using Vector 2022 skin with full content width enabled. This issue also occurs when viewing in my normal browser window, where I have the Vector 2010 legacy skin applied. I have also tested this in Firefox (version 115.0.2), and the same issue happens there as well. It even happens with the vector 2022 skin set to limited width mode.
My main display does have 125% scaling enabled, but if I drag the browser window to my secondary display which is set to 100% scale, the issue remains there also. — AP 499D25(talk)13:09, 30 July 2023 (UTC)
I don't know what exactly happens in the background, but MediaWiki considers {{coord}} to be part of the prose. Since page preview is supposed to show the first paragraph, it thinks that coord is that first paragraph, and shows it, but as it produces no text output, all you see is a blank. Similarly, in mobile view the first para should always be visible on top, then infobox, then everything else. So, coords above the infobox causes the software to think that coord is the first para, and the actual first para to be the second para, leading to the infobox at the top. But at the same time, please see MOS:ORDER which asks coords to be in the infobox or to be put near the end above categories. Putting coords above infobox is a popular but unsupported practice. —CX Zoom[he/him](let's talk • {C•X})20:09, 31 July 2023 (UTC)
Global Lua modules and Synchronizer tool
Hi! I figured some Lua developers may appreciate to know about this guide on how to develop a global Lua module (a module that can be used unchanged in all wikis) as well as the Synchronizer tool to automatically copy the master version of a module to all other wikis (and other neat features to make life easier).
I find both resources invaluable to maintain the modules I develop (Module:Excerpt and Module:Transcluder). After some initial hard work, I'm now able to update all wikis with a single click every time I push a new version live, as well as monitor any "forks" and fix them. Hopefully you'll find some of this useful! Sophivorus (talk) 23:24, 31 July 2023 (UTC)
Tech News: 2023-31
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Recent changes
The Synchronizer tool is now available to keep Lua modules synced across Wikimedia wikis, along with updated documentation to develop global Lua modules and templates.
The tag filter on Special:NewPages and revision history pages can now be inverted. For example, you can hide edits that were made using an automated tool. [6][7]
The Wikipedia ChatGPT plugin experiment can now be used by ChatGPT users who can use plugins. You can participate in a video call if you want to talk about this experiment or similar work. [8]
Problems
It was not possible to generate a PDF for pages with non-Latin characters in the title, for the last two weeks. This has now been fixed. [9]
Changes later this week
The new version of MediaWiki will be on test wikis and MediaWiki.org from 1 August. It will be on non-Wikipedia wikis and some Wikipedias from 2 August. It will be on all wikis from 3 August (calendar).
Starting on Tuesday, a new set of Wikipedias will get "Add a link" (Georgian Wikipedia, Kara-Kalpak Wikipedia, Kabyle Wikipedia, Kabardian Wikipedia, Kabiyè Wikipedia, Kikuyu Wikipedia, Kazakh Wikipedia, Khmer Wikipedia, Kannada Wikipedia, Kashmiri Wikipedia, Colognian Wikipedia, Kurdish Wikipedia, Cornish Wikipedia). This is part of the progressive deployment of this tool to more Wikipedias. The communities can configure how this feature works locally. [10]
@MBH, modifying "%{%{%ВАРБ:строка| at the beginning of the pattern to "%{%{%ВАРБ:(старая )?строка|, and changing line 262 from for candidate, to for _, candidate, should work. Please test! Izno (talk) 23:52, 25 July 2023 (UTC)
I don't understand OP's actual question, but I gotta wonder: Why does the pattern begin with %{%{%В and end with %}%}? If I put this in the debug console:
If I change that test to: =mw.ustring.match ("{{ВАРБ:строка", "{{ВАРБ:строка")
it returns {{ВАРБ:строка
I get similar results if I change mw.ustring.match() to string.match().
So, I guess that the first thing that Editor MBH must do is make sure that the pattern in ru:Module:Ballot#L-259 is a valid pattern. Remember that Lua patterns are not regex.
For further pattern consideration: what about whitespace around the template pipes and the '=' signs?
Ok. Don't believe everything that the debug console says. I hacked a test in my sandbox: Module:Sandbox/trappist the monk/ru test which shows that the pattern, even though flawed, works. Edit my sandbox and in the debug console type: =p.main() and press enter.
Surely, the author or any of the [several] technically experienced ruwiki editors can/should fix the ru.wiki module. Have you asked them?
I have asked them, they can't (they say, Lua doesn't support traditional regexes and they doesn't know Lua patterns) or doesn't have time for that. MBH (talk) 11:08, 27 July 2023 (UTC)
Aren't your debug-console inputs backwards? I tested with this:
...and got correct results in the debug console. The use of %В seems to be born of the confusion that, because %b is an actual lua pattern-matching escape, %В has to be escaped. Which of course is backwards logic anyway, but especially unnecessary because 'В' (the Cyrillic letter) isn't even the same as 'B', never mind 'b'. Still, the third %, while unnecessary, doesn't seem to harm anything.
Anyway, MBH, as far as I can tell mw.ustring.match doesn't support making entire parenthesized expressions optional (i.e. "%{%{ВАРБ:(старая) строка" will match "{{ВАРБ:старая строка", but "%{%{ВАРБ:(старая)? строка" will NOT), and it doesn't support alternation ("%{%{ВАРБ:(старая строка|строка)") so your only recourse is to have two different matches:
It isn't perfect because it is possible to write {{ВАРБ:some sort of nonsense строка|...}} which may not be desirable. If there can only be {{ВАРБ:строка|...}} and {{ВАРБ:старая строка|...}} then it should be safe...
Fixing the problem by rewriting the code to allow for whitespace, non-vertical format, parameters in any order should be done. To restrict ru.wiki editors in this way is poor programming practice.
^ Have to agree with this, too. Using matching expressions to essentially parse template arguments is... well, just icky, TBH. I don't know what this is for, but there's gotta be a better way. Even true regular expressions do not a parser make. Remember the immortal words of bobince at stackoverflow:
You can't parse [X]HTML with regex. Because HTML can't be parsed by regex. Regex is not a tool that can be used to correctly parse HTML. As I have answered in HTML-and-regex questions here so many times before, the use of regex will not allow you to consume HTML. Regular expressions are a tool that is insufficiently sophisticated to understand the constructs employed by HTML. HTML is not a regular language and hence cannot be parsed by regular expressions. Regex queries are not equipped to break down HTML into its meaningful parts. so many times but it is not getting to me. Even enhanced irregular regular expressions as used by Perl are not up to the task of parsing HTML. You will never make me crack. HTML is a language of sufficient complexity that it cannot be parsed by regular expressions. Even Jon Skeet cannot parse HTML using regular expressions. Every time you attempt to parse HTML with regular expressions, the unholy child weeps the blood of virgins, and Russian hackers pwn your webapp. Parsing HTML with regex summons tainted souls into the realm of the living. HTML and regex go together like love, marriage, and ritual infanticide. The <center> cannot hold it is too late. The force of regex and HTML together in the same conceptual space will destroy your mind like so much watery putty. If you parse HTML with regex you are giving in to Them and their blasphemous ways which doom us all to inhuman toil for the One whose Name cannot be expressed in the Basic Multilingual Plane, he comes. HTML-plus-regexp will liquify the nerves of the sentient whilst you observe, your psyche withering in the onslaught of horror. Rege̿̔̉x-based HTML parsers are the cancer that is killing StackOverflow it is too late it is too late we cannot be saved the transgression of a chi͡ld ensures regex will consume all living tissue (except for HTML which it cannot, as previously prophesied) dear lord help us how can anyone survive this scourge using regex to parse HTML has doomed humanity to an eternity of dread torture and security holes using regex as a tool to process HTML establishes a breach between this world and the dread realm of c͒ͪo͛ͫrrupt entities (like SGML entities, but more corrupt) a mere glimpse of the world of regex parsers for HTML will instantly transport a programmer's consciousness into a world of ceaseless screaming, he comes, the pestilent slithy regex-infection will devour your HTML parser, application and existence for all time like Visual Basic only worse he comes he comes do not fight he com̡e̶s, ̕h̵is un̨ho͞ly radiańcé destro҉ying all enli̍̈́̂̈́ghtenment, HTML tags lea͠ki̧n͘g fr̶ǫm ̡yo͟ur eye͢s̸ ̛l̕ik͏e liquid pain, the song of re̸gular expression parsing will extinguish the voices of mortal man from the sphere I can see it can you see ̲͚̖͔̙î̩́t̲͎̩̱͔́̋̀ it is beautiful the final snuffing of the lies of Man ALL IS LOŚ͖̩͇̗̪̏̈́T ALL IS LOST the pon̷y he comes he c̶̮omes he comes the ichor permeates all MY FACE MY FACE ᵒh god no NO NOO̼OO NΘ stop the an*̶͑̾̾̅ͫ͏̙̤g͇̫͛͆̾ͫ̑͆l͖͉̗̩̳̟̍ͫͥͨe̠̅s ͎a̧͈͖r̽̾̈́͒͑e not rè̑ͧ̌aͨl̘̝̙̃ͤ͂̾̆ ZA̡͊͠͝LGΌ ISͮ̂҉̯͈͕̹̘̱ TO͇̹̺ͅƝ̴ȳ̳ TH̘Ë͖́̉ ͠P̯͍̭O̚N̐Y̡ H̸̡̪̯ͨ͊̽̅̾̎Ȩ̬̩̾͛ͪ̈́̀́͘ ̶̧̨̱̹̭̯ͧ̾ͬC̷̙̲̝͖ͭ̏ͥͮ͟Oͮ͏̮̪̝͍M̲̖͊̒ͪͩͬ̚̚͜Ȇ̴̟̟͙̞ͩ͌͝S̨̥̫͎̭ͯ̿̔̀ͅ
Have you tried using an XML parser instead?
(Heh. The Russian hackers pwn your webapp. line is either irony or offensive, in this particular context.)
This module is for parsing pages with regular structure, consisting of substituted templates, these pages are created only twice per year and their contents are closely monitored (these are pages for the ArbCom elections). It is possible and even simple to parse templates or HTML with regexes if you know document structure beforehand. It's impossible only if you doesn't know document structure beforehand (for example, if you don't know how many nested elements there will be). MBH (talk) 20:12, 28 July 2023 (UTC)
Mmmm, that's true to a point, though it's kind of the whole point of {{subst:}} is that it dumps a template's output into an editable content page — creating a starting point, rather than an immutable end result, IOW. "Nobody should ever touch these pages because hand-editing their contents risks breaking the incredibly fragile module that extracts data from the template transclusions they hold" is the sort of onerous restriction Trappist was talking about. But I'll concede that it's far less bad when we're talking about purely behind-the-scenes administrative housekeeping stuff, rather than mainspace article content (or anything involved in creating article content), absolutely. FeRDNYC (talk) 23:06, 28 July 2023 (UTC)
Yeah, I'm looking at the code now... the problem, I think, is that you can't use an or conditional on the match execution itself, especially in a for loop like that. IOW, the if / or logic I suggested, incorporated into the module code, would look something more like this:
The truth tests on whether various patterns match anything at all (if they don't, they'll return nil which is a falsey value) have to be performed and a pattern chosen before using the for loop to extract the various subpattern matches. (The return {} in the else condition may or may not make sense, it's meant as a short-circuit to bail on the function if neither pattern matches the input string at all, since there are no values you can extract in those cases.) FeRDNYC (talk) 00:48, 28 July 2023 (UTC)
Oh, and BTW this code doesn't make any allowances for situations where both patterns can match the input string. If that's a possibility, then... well, then things get a lot messier. You may be better off rewriting the function so that it takes a pattern string as an additional argument. Then you can run it twice from the call site, once with each pattern. FeRDNYC (talk) 01:51, 28 July 2023 (UTC)
(Or, use Trappist the Monk's suggestions for building a single pattern that will work for both templates, even if it's not perfectly restricted to only those exact two strings.) FeRDNYC (talk) 02:28, 28 July 2023 (UTC)
Oh, the true reason why it doesn't work is because "старая строка" template has another parameter names, sorry for carelessness. I will try to integrate a different structure of second template into existing module. MBH (talk) 19:45, 28 July 2023 (UTC)
Ah! Well, then it might be simplest to just copy the function, and make a second version that processes transclusions of ru:ВАРБ:старая строка with its argument structure. If the arguments are different between the two, there's no longer any real advantage to processing them both inside the same function.
The other thing I'm noticing, even with the existing function, is: The code is extracting parameters it doesn't even use. Why? It's be more flexible to deal only with the parameters it actually needs, and ignore the others. For example, the code to find transclusions, and then to extract data from a given one, could perform those steps in separate stages. First, identify a transclusion and extract its parameter text as one big string. Then, parse each parameter out of that string, separately, so it's no longer critical where in the string each pattern is located. For instance:
Now your parameters can be in any order, they can contain arbitrary whitespace, it doesn't matter if they're newline- or space-separated, etc, etc. FeRDNYC (talk) 23:43, 28 July 2023 (UTC)
(Actually, if you use e.g. candidate=transclusion:match(candidate_pat), then you don't have to explicitly execute the function returned from :gmatch in order to get string results from it.) FeRDNYC (talk) 00:19, 29 July 2023 (UTC)
I'll take a look, sure. I would not be surprised if there's a function in the mw. package for performing the equivalent of all-pages-starting-with-prefix searches directly in Lua. (Or, barring that, we can probably make an API call and get the results back as JSON data.) FeRDNYC (talk) 06:03, 30 July 2023 (UTC)
Thanks, Izno, was returning here to report the same findings. So it doesn't look like that's a viable option, MBH, since there are requests to add such functionality, but they're still unfulfilled currently.
I had actually thought of a third possibility, having the two templates in question add the pages they're called on to a certain category. But that hits the exact same problem: the ability to retrieve a list of category members from Lua is T199126, also still unresolved.
In order to get the data to Lua in an importable form, I think you'd need to have a bot task that regularly re-scanned the subpages list and dump the results into a data file, which could then be imported with mw.ext.data.get. But it sounds like you've sorted things out, so that's good to hear. FeRDNYC (talk) 13:19, 1 August 2023 (UTC)
Not sure if this is the right place to ask, but I’ve noticed that when curly quotes and apostrophes (‘’ “”) appear in article space, bots like AutoWikiBrowser and Citation bot will more often than not “clean them up” and replace them with straight quotes, even when they are curly in things like citation titles. As far as I know curly quotes and straight quotes have the same exact purpose, so why exactly is more than one bot programmed to make every quote straight? 2001:4453:546:D000:C5F5:E244:E9A6:C47A (talk) 02:45, 2 August 2023 (UTC)
There is a FAQ at the top of WT:Manual of Style which has a short statement. There is also a "Search archives" box on that same page where entering "curly" will show a hundred arguments about the issue. Johnuniq (talk) 03:01, 2 August 2023 (UTC)
Cleanup after move bot
It's been a long time since I've seen it, but isn't there a bot that comes around and fixes redirects after moves? Is it still operating? Abductive (reasoning)03:39, 31 July 2023 (UTC)
Somebody (correctly) moved Echinopsis arachnacantha, a stub I created, to Lobivia arachnacantha. A couple of days later I happened upon it, and had to manually retarget a handful of synonyms. As far as my limited understanding of "double redirect" goes, these were double redirects. So I thought I'd ask here if the bot was still operating. Abductive (reasoning)08:02, 2 August 2023 (UTC)
There are some pages that should have TOC entries so that navigation could be easier.
First comes the module namespace. I often find it frustrating that there is no TOC entry to jump directly to the module code, the closest to it is the last entry of module documentation TOC, which unfortunately lands you 6-7 lines above your intended destination which is not good. Further, with the new sticky TOC in V-22, the last TOC entry remains bolded for the entire duration you look at the module code, which again is not good.
Secondly, Diff pages should have a TOC entry to separate the diff from the rest of the article, which is absolutely necessary when dealing with diffs with large-scale changes. Currently, the (Top) button of TOC will take you to the extreme top the page, and the next entry on TOC is the first header of the article. There is no navigational support to go directly to the beginning of the article prose (lead section), which again is not good for navigation because either you have scroll down a huge distance from the top, or scroll up some 3-4 paragraphs from the first header in order to reach the lead.
I am currently building a tool to check articles without images or with low resolution images (VCAT). I am currently using MediaWiki Action API action raw to get images avoiding images from templates and other non relevant images (example here). But I am having some issues with the Template:Infobox river, the image comes directly from Wikidata and does not appear in the raw of the article, like in the article the article Pur (Russia). (example of the tool not detecting images at: VCAT - Wikiproject Rivers).
Thanks, but I am still having issues, I am trying to identify articles without images, but in this way also non relevant images are included (eg. images in templates). For example in the article computational neuroscience there is no image, but using mw:API:Images it shows five images (example here) which are images used in templates. MingoBerlingo (talk) 08:05, 3 August 2023 (UTC)
Images from Wikidata in {{Infobox river}} are also images loaded through templates. You're basically saying you want images loaded through templates except the ones you don't want, and expecting the server to divinely know which ones are which. You could of course use prop=wbentityusage and detect images coming from Wikidata, but at that point I assume Quarry would be faster. Nardog (talk) 08:58, 3 August 2023 (UTC)
FYI, something like https://en.wikipedia.org/wiki/water?action=raw is no using the Action API. The equivalent via the Action API would be something like https://en.wikipedia.org/w/api.php?action=query&titles=water&prop=revisions&rvprop=content&rvslots=*. You can fetch images used in a page using mw:API:Images as pointed out above, but as you noted that gives all images while you only want certain ones. There's no way to directly get "just some images but not ones via some templates". If that's the sort of thing you want, you'd probably do best to process the HTML (either scraped or from mw:API:Parse) while excluding images inside elements with class metadata and the like. Another option would be to take the wikitext, strip out templates you don't like, and use mw:API:Parse with prop=images to get the list of images in what remains. Anomie⚔11:46, 3 August 2023 (UTC)
Would it be possible to generate an aggregate timecard for the UnderArmourKid LTA sockpuppets? I'd like to see what their pattern of behavior is over the last 90-180 days. I'm not sure if aggregating multiple users' edits into one timecard is already supported. ☆ Bri (talk) 17:51, 31 July 2023 (UTC)
@Certes: thanks for the suggestion, I coded up a Quarry query that's good enough. I'd still like those sweet graphics, though. ☆ Bri (talk) 19:41, 1 August 2023 (UTC)
Lately I've noticed that the move protection templates (e.g. pp-move) – and their small versions – are not rendering, at all. Example article being Jurassic Park. I recall these templates clearly working a few years back (at least back in 2017). I scoured through the talk pages for the templates (ahhh – something which I should've done for the article preview thread I created above beforehand), and I didn't find/see any recent threads talking about the lock not showing or rendering. The most recent thread I could find is from 2017 – this one – which just talks about it placing pages in the incorrect category, so it isn't relevant here.
I've tested the pages on both a normal browser window (where I have Vector 2010 skin selected) and in an incognito window, as well as in a different web browser, and it's the same result everywhere. Lock / banner not showing at all.
This appears to be intentional rather than a bug. Looking at Module:Protection banner/config (specifically line 860 onwards), it appears {{pp-move}} and similar have catonly set to yes by default. If you explicitly set |catonly=no, the templates will generate a banner/padlock about move protection. Aidan9382(talk)08:38, 1 August 2023 (UTC)
The template documentation states only appears for autoconfirmed users for when using the small version of the template. But I am autoconfirmed, yet it's still not showing at all. Furthermore, it only states that for the small version, it doesn't say it for the banner. I have personally never seen the full banner version of the template work before, in fact with pretty much every protection template I've seen, the full banner version only works less than half of the time. Btw I just came across a note on the top of the template talk page stating that it was nominated for deletion in October 2020, and also providing this link: Wikipedia:Templates for discussion/Log/2020 October 25#Template:Pp-move. So maybe looks like they did modify it to not show up for non-registered users, which makes a bit of sense anyways since anonymous users will never be able to move pages. Now, I just don't understand why it's not showing up for confirmed / autoconfirmed registered users like me. — AP 499D25(talk)08:52, 1 August 2023 (UTC)
The template documentation states only appears for autoconfirmed users for when using the small version of the template - {{pp-move}} has stopped producing a banner/padlock at all by default since july 2021, and the template documentation was just never updated to mention this behaviour.
in fact with pretty much every protection template I've seen, the full banner version only works less than half of the time - What do you mean? Looking at my own sandbox, the banners appear to work fine. They aren't often seen on articles since the standard appears to be to use |small=yes which makes them display as a padlock instead of a banner. Aidan9382(talk)09:06, 1 August 2023 (UTC)
So the template has been modified to simply not display a banner / lock / a visual indicator of any sort at all? If so, then that's pretty annoying that we have to click on a button and scan through a bit of the page source code to find out if a page is move-protected or not. I think the best way to handle it would be to have the template not show up for IP/unregistered editors, only registered users (maybe at the autoconfirmed and all above levels?).
As for the full banner protection templates not working half of the time, it's just something I swear I've seen a lot. Thinking more about it, one page that came up in my mind where it has a non-small protection template (pp-pc instead of pp-pc|small=yes), is Pop-Tarts. It does not display the full banner, just a small padlock like you would expect/get with small=yes. — AP 499D25(talk)09:23, 1 August 2023 (UTC)
The presence (or absence) of {{pp-move}} in the page source is not a guarantee that the page is move protected (or not). There are various methods of finding out the protection status, but the one that is both guaranteed and also requires no special privileges, no change to user preferences, no knowledge of coding and also the fewest clicks is to use the "Page information" link in the sidebar. Scroll down that to the "Page protection" heading, and it's all in the table. --Redrose64 🌹 (talk) 12:48, 1 August 2023 (UTC)
That makes sense, I'm aware of that. Though I didn't know about the protection status in the "Page information" view, so thanks for telling me about that!
Still though, the move protection would be an instant and easy indicator as to whether a page is move protected or not (i.e. no need to click a link to find out). Like I said above, I think it should be visible to registered users only. Don't we have bots that automatically add or remove protection templates when a page is protected or unprotected?
I'm going to guess the use of move-protection templates as a "visual indicator" have been deprecated and they are just purely used for page categorisation purposes now. — AP 499D25(talk)02:15, 4 August 2023 (UTC)
Can't move text on userpage
I'm not sure if there is another place to ask, but this seems to be some kind of bug.
I was on my user page when using Chrome and saw that a quote I put there looked very strange, so I tried multiple times to fix it. It is almost fixed now but no matter what I do, the words "The file is deceased" end up on the wrong line. If the quote isn't there, there's just this big empty space.— Vchimpanzee • talk • contributions • 22:48, 2 August 2023 (UTC)
How about now? A line break in wikitext converts to space, so you have to use <br> or <poem>...</poem> if you want to realize a line break in output. Nardog (talk) 23:16, 2 August 2023 (UTC)
I could (if someone would explain how) but it wouldn't be helpful to anything important. I don't think that quote is of any value on my userpage anyway. I liked it when I saw it but I know things looked right when I took it out.— Vchimpanzee • talk • contributions • 15:39, 4 August 2023 (UTC)
I have often found that HotCat randomly fails when making single quick changes. However, I have just found that it consistently fails to commit a change that involves removing a blank sortkey. –LaundryPizza03 (dc̄) 16:38, 4 August 2023 (UTC)
I assume CAPTCHA has been turned on for all IP editors for all edits, or something happened recently needing CAPTCHAs. I was wondering why a CAPTCHA is required for previews? Shouldn't that not require CAPTCHA, as repeatedly entering it for the same edit, might end up dissuading people from contributing at all. Or are the servers overloaded, so that every preview would need a CAPTCHA? -- 67.70.25.80 (talk) 06:42, 2 August 2023 (UTC)
An emergency measure to require extra CAPTCHAs has successfully prevented disruption and has just been reverted, according to the ANI thread. Certes (talk) 10:51, 2 August 2023 (UTC)
I was about to start my own thread about this. I've been having to enter in CAPTCHA for every single edit I've made now. Before, I believe CAPTCHA was only required for edits that add new external links to pages. Is this a new policy? 2601:1C0:4401:F60:E07A:CA49:686C:2E64 (talk) 19:25, 2 August 2023 (UTC)
Agree that it makes no sense to fill in a CAPTCHA for a preview. And it appears that this is only enforced on the client side. When I enter random text in the captcha field, it does show the preview without any hiccups. 93.72.49.123 (talk) 21:36, 2 August 2023 (UTC)
Unless something changed in the last few hours, I don't think so. I made a slight spelling correction while logged out and didn't need to enter a captcha. WPscattert/c19:31, 2 August 2023 (UTC)
Oh wow, something did change in the last few hours. How about that. Oddly enough I was logged out by complete accident when I made the edit this morning, and I think it's my only logged-out edit since I created an account. Here I thought how serendipitous that I could answer this question with recent personal experience! Thanks for the info! WPscattert/c19:42, 2 August 2023 (UTC)
In the meantime, I filed task T343585 about the fact that filling out the field is required when previewing. It should only be required when publishing your changes. Matma Rextalk18:26, 4 August 2023 (UTC)
The "Page preview" feature shows gibberish with no clear origin for links to one particular page
With the Page previews feature enabled, hovering the word and hyperlink to Entente Cordiale, the snippet reads: "{{sidemianonn dominga michele daniel cosma enea" (curly braces included, open only; quotation marks not included).
Observed in In [[Concorde#naming] but it also appears in the preview to this very post.
Vector Legacy and Vector 2022 CSS cross-contamination
I am trying to use different custom css for Vector Legacy and Vector 2022, for the purposes of testing dark mode features, and am finding that the css specified for vector Legacy (vector.css) is loading even when using vector 2022 setting in preferences and the vector-2022.css. Is this intended behavior? is there any way to specify css that will only load when using vector legacy and not load when using vector 2022? Dialectric (talk) 17:08, 4 August 2023 (UTC)
It was intended behavior, and sometime toward the end of the year that intended behavior will no longer be intended and will be turned off. (On the list of things to RFC is to do it earlier than everyone else and be done with it.)
You may target specific skins, the classes for which are set on the body element. I think one is skin-vector-legacy and the other skin-vector-2022 off the cuff. Izno (talk) 17:43, 4 August 2023 (UTC)
Thanks for the reply. I did target some css with .skin-vector-2022; looks like .skin-vector-legacy will do for limiting some css to the legacy view.Dialectric (talk) 18:09, 4 August 2023 (UTC)
The <body> tag has several classes in its class= attribute. As I write this, those for Vector (legacy) include skin-vector and skin-vector-legacy, those for Vector-2022 include skin-vector and skin-vector-2022. --Redrose64 🌹 (talk) 20:30, 4 August 2023 (UTC)
Note that you can use CSS nesting which is now available in most browsers (also in Firefox behind a `layout.css.nesting.enabled` flag). You will get some warnings from MW linter, but it seems to work fine. Nux (talk) 13:56, 5 August 2023 (UTC)
If you do something weird with & or do many, many levels of nesting, then sure, it might stop working. But it is already implemented in Safari and Chrome and will not go away [13]. Just adding `.skin-vector-2022` around other CSS rules will definitely work in the future. Nux (talk) 19:56, 5 August 2023 (UTC)
Rules that apply only to Vector 2022 should be placed only in your Vector 2022 CSS page, which means that CSS needs no wrapping. It is solely CSS in Vector.css that needs clarification as to application. Izno (talk) 20:10, 5 August 2023 (UTC)
"failed to parse" errors
At the moment the search function finds 38 cases of the following error:
Failed to parse (SVG (MathML can be enabled via browser plugin): Invalid response ("Math extension cannot connect to Restbase.") from server "http://localhost:6011/en.wikipedia.org/v1/":): f
"Re-review this revision" – footer box about very old edit
Viewing the page Communication and scrolling to the bottom of the page, I see a bordered box asking me to re-review a two-year old edit, and linking to a Feb. 2021 log item by Samsara, inviting me to "Accept" or "Unaccept" it:
This log action occurred more than 500 edits ago at the article, and I don't know why that box is appearing now (or at all). Mathglot (talk) 19:18, 4 August 2023 (UTC)
It's not asking you to re-review a two-year old edit, it's just showing the log entry of when the page was pending-changes protected so you can see the context or rationale for the protection. The pasted source shows the revision being reviewed is 1167651378, which was the last revision when you created this thread. Nardog (talk) 17:18, 6 August 2023 (UTC)
I cannot translate page from polish to english
Hello!
I wanted to translate page about Nastazja Staniszewska from polish to english, but wikipedia says that it could be done only by experienced users. What can i do about it? MadAurochs (talk) 15:12, 7 August 2023 (UTC)
If this is the Polish-language Wikipedia biography you are referring to, [14] I suspect a direct translation wouldn't be accepted, since by English-language Wikipedia rules, more coverage in independent sources would probably be needed to meet our notability criteria. You'd need to find such sources, and incorporate them into the article.
You cannot use WP:CX if you are not WP:Extended confirmed, meaning you need to have been on English Wikipedia for 500 edits and 30 days. If you still believe the biography is suitable, 1) you will need to do the translation by hand, and 2) I would recommend using WP:AFC for additional review. Izno (talk) 17:20, 7 August 2023 (UTC)
Tech News: 2023-32
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Recent changes
Mobile Web editors can now edit a whole page at once. To use this feature, turn on "⧼Mobile-frontend-mobile-option-amc⧽" in your settings and use the "Edit full page" button in the "More" menu. [15]
The drop-down templates in the Edit Window have always had a feature that if you click on one of the templates, input an URL, and clicked on the little autofill symbol next to it, it would automatically fill out the template. As of today, that's not working. I have three different browsers, and it's not working on any of them. It was working fine recently. Maybe it's because I'm trying to do that with a PDF? Feedback? — Maile (talk) 23:10, 7 August 2023 (UTC)
Template:Welcome-COI-acc already has logic (near the end) to suppress the category if the template was added to the page years ago. Perhaps it could also check that the page title doesn't end in ".js". Certes (talk) 19:00, 5 August 2023 (UTC)
The typical way to deal with this is simply to wrap the JS page in <nowiki>s. Since these are automatically changed, that could be fixed on the script side.
On the software side, while I have observed no actual consensus on the matter, my belief is that we want to continue to allow wikitext parsing in such content model pages. You could of course file a task to turn it off for JS content models (just as it is off for Scribunto). Izno (talk) 19:30, 5 August 2023 (UTC)
For the latter item, yes. For the former, that would be at WT:TW. I have observed some people say it is a good thing, not a bad thing, because that it is what allows such pages to be categorized, among other qualities. Izno (talk) 00:22, 6 August 2023 (UTC)
Before you request it be turned off, you might want to review T43155 where the parsing was explicitly re-added after it was accidentally removed back in 2012. It's one of those things that wouldn't be done if this were new code, but may need to be kept for hysterical raisins since it already exists. Anomie⚔01:42, 6 August 2023 (UTC)
Another benefit of parsing is that many .js comments include wikilinks to imported scripts, enabling toolsmiths to notify their customers of changes. Certes (talk) 17:56, 6 August 2023 (UTC)
What would probably be most lost is categorization. en.wp doesn't tend to categorize their scripts but I think other wikis might. Izno (talk) 21:43, 6 August 2023 (UTC)
One thing to potentially experiment with is whether JSON content model pages do the same as JS pages in parsing stored wikitext. That could be a viable alternative for storing the config. Izno (talk) 17:24, 6 August 2023 (UTC)
Looks like twinkleoptions.js is already wrapped in nowiki. Example. The diff in the original post, User:AfroThundr3007730/twinkleoptions.js, is from 2018, so perhaps that is from before the nowiki patch was deployed to Twinkle. Looks like User:AfroThundr3007730/twinkleoptions.js is the only twinkleoptions page in that particular category. Search. I'll go ahead and use my interface administrator permissions to fix the original page. If there's a low number of other pages discovered, let me know and I can do some manual editing. –Novem Linguae (talk) 08:41, 8 August 2023 (UTC)
The SVG image creators create font family lists with a choice of fonts that meet the demands of Windows, Macs, Linux, and even Wikimedia.
So if Wikimedia just left it alone it would work fine. As long as there was a free font at the end of the font family list that was metric-compatible to the other fonts. See :
Wikimedia has the narrow free font Liberation Sans. But it is not being used. A wider font is being substituted for one of the other fonts in the list. Possibly DejaVu Sans:
I decided to make a test: It appears that it either isn't processing fallback fonts anymore (just jumping right to the default Deja Vu Sans) or every unknown font is being "defined" as Deja Vu Sans so fallback never happens. Either way that sounds like a bug that should be reported in Phabricator. Anomie⚔22:06, 8 August 2023 (UTC)
Thanks, Anomie. Maybe you can post this at Phabricator. I am almost always baffled by aspects of it when I post there. You sound like you are more familiar with it.
From your image it looks like DejaVu Sans is being substituted for Lato and 'Helvetica Neue' and bogus font names. It looks like Liberation Sans is being substituted for Helvetica and Arial.
I think your image test would be better if "This is a test:" was in front of them all. Then comparisons would be longer and clearer. Right now I am just comparing "This is" for each one.
It looks like the narrower Liberation Sans should be the default fallback font instead of the wider DejaVu Sans characters and character spacing. --Timeshifter (talk) 22:52, 8 August 2023 (UTC)
Vector 2022 table of contents: bolded heading differing from current scroll position
Usually the bolded heading in the table of contents on the side with the Vector 2022 skin matches the current scroll position in the primary page content area. However at present, when viewing the incidents' noticeboard, this behaviour only persists when the current heading or sub-heading is near the top of the page. Once it scrolls away, the bolded heading reverts to a heading of a different section. This behaviour seems to have started with this edit to the section in question. The preceding version behaves as usual, while the following version does not. Can anyone figure out how the wikitext included (via transclusion) to collapse some headings has caused the change in behaviour, and how it may be fixed? (Pinging SGrabarczuk (WMF) and OVasileva (WMF).) isaacl (talk) 18:44, 5 August 2023 (UTC)
I have a question about the abuse filter: How can I write a filter to block users who continuously add spam content like the example in the image on the right-hand side?
Would need to target something. Got any ideas? Are the accounts all new editors? Do they put the same thing in each of their edits? Do they put the same thing in each of their edit summaries? Do they only target certain pages? etc. –Novem Linguae (talk) 06:48, 6 August 2023 (UTC)
Both accounts that I mentioned above are new accounts. Shortly after their creation, they started engaging in disruptive behavior. As for other factors, I will investigate further. Thank you. Plantaest (talk) 06:52, 6 August 2023 (UTC)
For an simple filter you could do 'contains_any(added_lines, "target line")' where 'target line' is whatever words he is adding. When you find one, I would use the search to see whether other people are using that same phrase. You can use vi:Special:AbuseFilter/test within 30 days of the edits to test what your filter is matching. Snævar (talk) 19:31, 8 August 2023 (UTC)
@Plantaest You're going to need to find "common patterns" on your own but something like below?
In any case it's neccesary to debug your code over and over to eliminate false positives before enabling the filter, especially when you choose block as the action to take. Dragoniez (talk)03:26, 9 August 2023 (UTC)
Detect multiple transclusions of a template within a page
I'm looking for a way to detect multiple transclusions of a template within a page using a module, to add a maintenance category. Let's say we have {{A}} {{A}} in a single page, in which case I want to add e.g. Category:Multiple transclusions, but none when we only have {{A}} (single transclusion). Does anyone have any ideas? Dragoniez (talk)02:57, 9 August 2023 (UTC)
This is not information that is tracked at a database level. Why do you want this ? Generally multiple transclusions shouldn't really be too much of a problem. —TheDJ (talk • contribs) 10:05, 9 August 2023 (UTC)
Many clever people have tried to answer similar questions here, and I don't think it is possible. Would the module which adds the category be called from template A or in some other way, or is either ok? If you tell us more about the problem, we may be able to think more laterally and find an indirect solution. Certes (talk) 11:08, 9 August 2023 (UTC)
Say, A is a template like {{Old XfD multi}}, which can have a bunch of "data rows". These rows don't need to be separated into multiple occurrences of the template and can be merged. I'm wondering if a bot can do this merging task. But if you develop such a bot, you'll need a way to get a list of the relevant pages, and I thought it'd be one option to have a tracking category so that we can collect page titles using list=categorymembers. I can think of other ways to do this, like using list=search, but the category solution seems to me to be most efficient, although I don't know if it's fundamentally possible to get a lua module to add such a category. This is why I thought I'd ask this. Dragoniez (talk)12:04, 9 August 2023 (UTC)
In a module can't you just use string.gmatch for the template name (or {{[Tt]emplate name) on the page content and iterate the result to get the count? — Jts1882 | talk16:53, 9 August 2023 (UTC)
So I did the only logical thing: Make a JavaScript program to get all the base domains from the table, and then (through a CORS proxy) see which ones are broken.
@QuickQuokka I'm not sure if you're looking for an answer to a question, or a pat on the back (LOL), but this sounds far too technical a point to raise at the Teahouse. I wonder if you might be better off making whatever point it is that you're making over at WP:VPT? I'm really not trying to be rude, but the folks over there understand much the same sort of stuff that you seem to. Me on the other hand: nah! Best wishes, Nick Moyes (talk) 20:30, 9 August 2023 (UTC)
That column has data-sort-type="number", so the entries are sorted numerically. Entries with no numeric value remain sorted by country name, regardless of what non-numeric value (Illegal, None, etc.) they contain. Certes (talk) 11:47, 10 August 2023 (UTC)
It seems like the fire diamond in Chembox Template for all chemical compounds is misaligned on mobile devices. The background image is moved completely to the right and some of the numbers / letters overlap the diamond's boundaries. I've checked on multiple devices - all with latest version of Chrome Browser. The alignment is alright on PC. Check example : Benzene. I can't fix it (as I have no idea how it works). Can anyone having knowledge about it help? Thanks in advance!! Ray Frost (talk) 18:18, 10 August 2023 (UTC)
All the diamonds are misaligned to the down-left. Is there a way to edit it? As I said, I have no idea how that code works and I really don't want make it worse. Ray Frost (talk) 09:07, 11 August 2023 (UTC)
This is because the <figure> around the image has large margins on mobile, while it has no margins on desktop. This might be an oversight in the mobile styling for images, but I'm not 100% sure. It can be corrected by making use of TemplateStyles to style this element. —TheDJ (talk • contribs) 11:04, 11 August 2023 (UTC)
@Jonesey95, .content is "outside the reach" of TemplateStyles. All TemplateStyles selectors are "hoisted" to .mw-parser-output. This means that when you wrote .content, what you really wrote was .mw-parser-output .content. Now, the .content element exists outside the .mw-parser-output element, so in reality you were targeting nothing, as when you have multiple selectors they must be in the order of the HTML containers. This is something specific to TemplateStyles and that selector would have otherwise worked if it were in, say, Common.css, since that does not hoist CSS. The hoisting is one of the reasons that TemplateStyles is "safe". (There is one exception to how this hoisting works: you may target classes on body and html if you write them as body.class or html.class so long as you pick another child element, say, body.class .infobox. The corresponding output will be body.class .mw-parser-output .infobox and similar.)
Hey, during a Quarry query today, I saw that there was a replag on the English Wikipedia. Then, coming here, I remember that when this happened last time, it was suggest that a link to the replag page be put in the VPT FAQs page but that doesn't seem to ever been done. Anyone want to take on that small task? LizRead!Talk!02:06, 12 August 2023 (UTC)
AutoEd and templates
It would be possible with AutoEd to create a configuration with a template and a list of parameters and have it remove the parameters in that template that are not in the list. --GryffindorD (talk) 10:47, 12 August 2023 (UTC)
Probably. I have an example at User:Jonesey95/AutoEd/unnamed.js. It fixes misnamed parameters, but the logic is similar. You may have to list the invalid parameters individually, which is probably a good idea for most templates, because people often misspell valid parameters, so those should be fixed instead of removed. – Jonesey95 (talk) 13:05, 12 August 2023 (UTC)
It is sometimes difficult to figure out which parameter creates the error, though, particularly with templates I know little about. If it were possible to get AutoEd to remove them, in previewing you would immediately understand the problem and possibly fix it. GryffindorD (talk) 13:17, 12 August 2023 (UTC)
Can't use MOS scripts
I'm using a few MOS scripts for copy-editing, the list is at User:Lallint/common.js. However, the scripts, for example the EngvarB scripts, only seem to work on my common.js page, and not in article namespaces. Going into source editor with the scripts in article namespace does show some of the scripts, but some are missing, and the ones that show do nothing when clicked. When I go into source editor on my common.js page, all of the scripts show, and I can use the scripts as intended. I'm using the latest version of Firefox, and I've tried chrome but it does the same thing. I've had this problem for a year, but its still happened. Lallint01:05, 12 August 2023 (UTC)
Those scripts are not compatible with the 2017 editor. Turn off "Use the wikitext mode inside the visual editor" in Preferences. Nardog (talk) 05:38, 12 August 2023 (UTC)
Could someone help out? I added colors to my signature and Wikipedia provided a notice saying
"Code that you insert on this page could contain malicious content capable of compromising your account. If you import a script from another page with "importScript", "mw.loader.load", "iusc", or "lusc", take note that this causes you to dynamically load a remote script, which could be changed by others. Editors are responsible for all edits and actions they perform, including by scripts. User scripts are not centrally supported and may malfunction or become inoperable due to software changes. A guide to help you find broken scripts is available. If you are unsure whether code you are adding to this page is safe, you can ask at the appropriate village pump.
This code will be executed when previewing this page."
@Ktkvtsh: That is the standard notice that appears on everyone's .js and .css pages. It's just telling you to be careful what you add, not that there's anything wrong with what you have now. gnu5713:11, 13 August 2023 (UTC)
As gnu said, the warning is generic; it just says that your computer will run anything written there, so if you copy and paste text from an untrusted source then it can do bad things. What you've added looks good, but beware that it will only affect how you see your signature when you are logged in. If you want your signature to appear differently to everyone, you need to change your preference in the Signature section of Special:Preferences#mw-prefsection-personal. Certes (talk) 13:14, 13 August 2023 (UTC)
Right now, if you visit an old revision or diff of a page you're watching, it marks the latest revision as read, i.e. removes all "updated since your last visit" labels on the page history. Has this always been the case? It should only mark as read the revisions up to that point. And I feel like it did until weeks or months ago, or am I misremembering? Nardog (talk) 11:09, 13 August 2023 (UTC)
Weno has a broken infobox displaying a giant redlink. I can't figure out how to fix it as it's somehow procedurally generated. Anyone know what to do? Ten Pound Hammer • (What did I screw up now?)17:40, 9 August 2023 (UTC)
The template documentation is no help; it just says it "should not be used from article namespace pages". (Didn't we have a whole bunch of giant rfcs all saying not to do the sort of things that template does? Or did they suddenly start going the other way over the last year while I've been mostly inactive?) —Cryptic17:49, 9 August 2023 (UTC)
It looks like the problem is that Template:Infobox settlement/Wikidata says {{#if: {{#property:P31}} | [[{{#property:P31}}]] }} which then strings the two P31 parameters of wikidata:Q1009384 together and turns them into a link. I am not sure that P31 can always be interpreted as "settlement type", certainly not as something that can be turned into a piped link. Jo-Jo Eumerus (talk) 18:16, 9 August 2023 (UTC)
Per the RFC, that Wikidata property call is invalid, since it does not check to see if the value is sourced. I have expanded the documentation for the template to explain that it should not be used in article space until it is modified to conform to the RFC's sourcing requirement. – Jonesey95 (talk) 18:25, 9 August 2023 (UTC)
I'm honestly inclined to just nominate it for deletion - the chance of the warning you added becoming anything other than dead letter otherwise is negligible and it seems like nobody wants to do the necessary work. * Pppery *it has begun...18:52, 9 August 2023 (UTC)
Does anyone know whether {{#invoke:WikidataIB|getValue|rank=best|qid={{{qid|}}}|Pxxx|fetchwikidata=ALL|onlysourced=yes}} is the right replacement for {{#if: {{#property:Pxxx}} | [[{{#property:Pxxx}}]] }} code? To solve the source thing. Jo-Jo Eumerus (talk) 20:01, 9 August 2023 (UTC)
It was my understanding that Google only lists Wikipedia articles that are in article space, also known as mainspace. I also thought that Google only listed articles that have been reviewed by New Page Patrol, but that is not my concern in this post. However, I have seen that a Google search for Zepotha shows Wikipedia:Zepotha. For background, Zepotha is a fictitious movie, apparently originating on TikTok. Multiple pages about it were recently deleted from draft space as a hoax. I tried to search Google for non-Wikipedia mentions of this name or meme. As my screen shot shows, Google displays Wikipedia:Zepotha, which is a red link showing that it has been deleted three times by administrators.
So my question is whether anyone is aware of a way that Google lists pages in Wikipedia project (WP) spaceRobert McClenon (talk) 05:16, 14 August 2023 (UTC)
User:Cullen328 - Thank you, sort of. In this case Google was stupid. Articles in spaces other than article space will seldom be of interest to a reader who is using a search engine to search the World Wide Web. Pages in project space are seldom of interest to someone who is not a Wikipedia volunteer. In this case it will just confuse the searcher. Duh. Robert McClenon (talk) 06:07, 14 August 2023 (UTC)
Robert McClenon, the only website that Wikipedia editors have any direct influence on is Wikipedia. As for Google, they are useful 98% of the time, but occasionally, they screw up. For example, their navigation instructions to my neighborhood tell people to drive through a locked gate to get to my house, and when a driver passes by the locked gate, the directions turn into complete nonsense. I have to warn guests. Cullen328 (talk) 06:16, 14 August 2023 (UTC)
I'm not sure why ppl think that only main space is of interest to Google or people trying to find information on Wikipedia for that matter. They might prioritise main space higher internally in their algorightme of course, but information is information. If you put it on the web, Google will index it and probably should index it. While we exclude some pages that might be used to game the system, or that are more likely to include personal information etc. In general, every single page is indexed. —TheDJ (talk • contribs) 13:36, 14 August 2023 (UTC)
How to Find??
Identify political leader pages lacking the Infobox officeholder template, along with a user-friendly filter for sorting them by election or state. I have work on Tamil Nadu state & 2021 Assembly Election. Thanks in advance. - IJohnKennady (talk) 03:36, 6 August 2023 (UTC)
Maybe try Petscan, with a category selected, a high depth of subcategories (e.g. 10), and a "doesn't have template" {{Infobox officeholder}} selected. Example. Perhaps choose more specific categories and templates to reduce the # of results to a manageable level. Hope this helps. –Novem Linguae (talk) 04:39, 6 August 2023 (UTC)
Help:Searching says the maximum search string is 300 characters, which matches my experience. However, it now seems to be accepting only 255 characters. Has it changed? This breaks some existing searches. (Per phab:T107947, the restriction seems to be an arbitrary one to discourage slow searches, but I'm having to remove optimisations to fit within the lower limit. Has someone scored an own goal?) Certes (talk) 09:38, 14 August 2023 (UTC)
Please be more specific. Generic Ajax rollback is now in core ("Show a confirmation prompt when clicking on a rollback link") so you don't have to import any script, and if you use a more elaborate script you most likely have to turn the option off. Nardog (talk) 09:06, 13 August 2023 (UTC)
I mean the functionality that used to be able to rollback on special:contribs, without redirecting to another page... ~Loftyabyss09:27, 13 August 2023 (UTC)
Blank your global.js, common.js, and the .js for whichever skin you're using, and turn on "Show a confirmation prompt when clicking on a rollback link". Nardog (talk) 18:21, 14 August 2023 (UTC)
But confirmations isn't what I'm seeking? There were these [ajax] links which did it with no confirmation or page redirections, once... ~Loftyabyss19:35, 14 August 2023 (UTC)
On what wiki have you seen it? Whatever it is, it's a user script or gadget, not something from the MediaWiki software or an extension. Nardog (talk) 02:52, 15 August 2023 (UTC)
Tech News: 2023-33
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Recent changes
The Content translation system is no longer using Youdao's machine translation service. The service was in place for several years, but due to no usage, and availability of alternatives, it was deprecated to reduce maintenance overheads. Other services which cover the same languages are still available. [18]
Changes later this week
The new version of MediaWiki will be on test wikis and MediaWiki.org from 15 August. It will be on non-Wikipedia wikis and some Wikipedias from 16 August. It will be on all wikis from 17 August (calendar).
Starting on Wednesday, a new set of Wikipedias will get "Add a link" (Latin Wikipedia, Ladino Wikipedia, Luxembourgish Wikipedia, Lak Wikipedia, Lezghian Wikipedia, Lingua Franca Nova Wikipedia, Ganda Wikipedia, Limburgish Wikipedia, Ligurian Wikipedia, Lombard Wikipedia, Lingala Wikipedia, Latgalian Wikipedia, Latvian Wikipedia, Maithili Wikipedia, Basa Banyumasan Wikipedia, Moksha Wikipedia, Malagasy Wikipedia, Armenian Wikipedia, Kyrgyz Wikipedia). This is part of the progressive deployment of this tool to more Wikipedias. The communities can configure how this feature works locally. [19]
Should work now. When you create a .css page in the user space, it is treated as a CSS content model. The reasons for that are somewhere in the "security" basket. An interface admin (such as myself) can change the content model to sanitized CSS. Izno (talk) 07:18, 15 August 2023 (UTC)
Take a look near the bottom of Berliner FC Dynamo, in the sections "References", "Further reading", and "External links". {{reflist}} and {{cite news}} templates are not displaying as intended, but rather as links to the template pages. I cannot see what is causing this. Your help would be appreciated. DuncanHill (talk) 10:45, 14 August 2023 (UTC)
Its got over 1100 references. Many sentences have multiple, so perhaps some can be cut down. The flags could also be cut, but that is only a small amount all up. Graeme Bartlett (talk) 11:55, 14 August 2023 (UTC)
I don't know where to lodge this information about Wikipedia:WhoWroteThat, since the page is a redirect to mediawiki; perhaps someone here is also active there or knows who to ask on en.wiki about this.
Until Bocaranda, I've found WhoWroteThat to be reliable. Lately, it is glitching at Nelson Bocaranda, and I'm wondering if the interlanguage links are somehow throwing it off (since I started noticing it about the time I added them, but I can't be certain that was it). Today, it is attributing a ton of text that I wrote to WMrapids; as examples, look at the sentence using "knee pain", and the sentence starting with "Contrary to Article 143". And then halfway through the "Responses" section, it stops working entirely. SandyGeorgia (Talk) 15:18, 10 August 2023 (UTC)
In the future, for WhoWroteThat, bugs can be filed directly on Phabricator using this link. Just make sure to put "Who-Wrote-That" in the tags section. This will skip a couple of middlemen and pings. Creating a ticket with the correct tag emails the developers of that tool, such as MusikAnimal. Hope this helps. –Novem Linguae (talk) 12:30, 11 August 2023 (UTC)
I appreciate that info Novem Linguae but I'm unlikely to file a report directly-- not familiar enough to know if something is truly a bug or just me. Instead, I am more likely to ping you directly :) :) SandyGeorgia (Talk) 15:48, 15 August 2023 (UTC)
Question about talk page blocks and protection
Hypothetical question I can't figure out: if an IP editor is blocked but has not had talk access revoked, and their user talk page is semiprotected, can they edit the page? It would be an unusual situation but I'm wondering what the actual answer is. Ivanvector (Talk/Edits) 16:26, 15 August 2023 (UTC)
For the past fortnight the text on the edit source page has frequently randomly shrunk so that only three lines are visible and can be edited. I have tried using a diferent browser and a different computer and neither solved the problem.
I have a screenshot but it is in Word and the instructions say it has to be png, so I would have to find out how to convert it if uploading it would be helpful. Thanks if anyone can help. Dudley Miles (talk) 18:44, 14 August 2023 (UTC)
first thing i'd suggest is to try "safemode" (add safemode=1 to browser address line, preceded with ? Or if it already contains ?, then with & ). If the issue goes away in safemode, it means most probably some script or gadget you have installed or activated is the culprit. If it doesn't go away, and presuming windows operating system, use snipping tool (windows key, type snipping tool, and click on the app when you see it), which lets you save as png to capture the screenshot.
A screenshot would be very helpful for visualizing the problem, since there was no steps to reproduce provided. To convert a screenshot from Word to PNG, try the following.
1) Open an image editing program such as MS Paint
2) Open your word Document
3) Right click on the image -> copy
4) Switch focus to Paint
5) Paste
6) Save as -> something.png
7) Upload to Wikimedia Commons or your preferred third party image host
So the edit box has its normal dimensions, but only three lines inside it can be used (as shown by the scrollbar on the right). I've not come across that before. --Redrose64 🌹 (talk) 13:18, 15 August 2023 (UTC)
I have followed MusikAnimal's instructions and the problem code appears to be "importScript('User:Lingzhi2/reviewsourcecheck-sb.js'); // Ling's source review script". It is curious that I installed the code in September 2021 and I have not had a problem until now, but I deleted it and trust this will solve the problem. Thanks for all the help. Dudley Miles (talk) 20:04, 15 August 2023 (UTC)
How to add an academic thesis in Wikidata
I stumbled upon Peter Osborne (philosopher), whose infobox provides his thesis title and URL. I went to his Wikidata page and attempted to add the academic thesis there. Compared to Wikipedia, this page has a strange interface. I found "add statement" at the bottom of the page. I clicked on that, then entered "academic thesis" in the first field, then I entered "The carnival of philosophy : philosophy, politics and science in Hegel and Marx" (with and without a trailing period) and "publish" is grayed out. Then I added 1 Reference with reference URL "https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.253269". "publish" is still grayed out. How do I get "publish" to be clickable? Do I have to first create the thesis in Wikidata as a separate entry before I can add it to this entry? —Anomalocaris (talk) 17:58, 15 August 2023 (UTC)
Xaosflux: Following those directions, after "put Q121477887 in there", there's ! in a circle next to the thesis name, and clicking on it shows "Values of academic thesis statements should be instances of one of the following classes (or of one of their subclasses), but The carnival of philosophy : philosophy, politics and science in Hegel and Marx currently isn't:" followed by a bulleted list that is variations on "thesis". Which is exactly what I said in my previous comment, the thesis entry is miscategorized as something other than a thesis and there's apparent way to to reclass it. —Anomalocaris (talk) 20:06, 15 August 2023 (UTC)
Unfortunately JD Forrester dismissed these with his "won't fix" comment "Eurgh, yeah, I don't think namespaceDupes will fix those. There are only three, low-value redirects on enwiki, but across the cluster there are likely a few that we'll be sad to lose." and Denny functions closed the task as Resolved.
So once again I find that if you really want to get something done, it's better to ask at the village pump rather than Phabricator (where you're most often dismissed or ignored).
@Taavi: can you work the same magic with f: you did with fat: using the page move API and page IDs queried from the replicas for "F:NV", "F: NV", and "F:F:F"? Thanks, wbm1058 (talk) 14:55, 15 August 2023 (UTC)
Contrary to your claim, I did not dismiss this concern, I complained that we didn't have a way of fixing it automatically, and expressed sadness that we can't fix it automatically, unlike when namespaces are added (because interwikis work in a different way).
You (or someone) will need to make a series of action=edit API calls to manually fix each of them; unlike pages hidden by namespace changes, they're still available via their page ID in the API.
Could someone report a bug for me? As an IP I don't have access to phabricator.
The issue is that the __NONEWSECTIONLINK__ does not properly supress the "new section" tab on certain "action" pages, and that it is still possible to add new sections to the page using discussion tools.
Strange one that. It's easy to make it display the list from 7 days ago instead of 8, but it looks like it's always been like that. Is it actually being used anywhere? Black Kite (talk)15:05, 11 August 2023 (UTC)
I think the reason I made it 8 days was because the AfD should be deleted or kept after seven days of discussion (it would be the eighth day when the AfDs would be closed). Just double check and make the change if you think that is prudent. Thanks, Lourdes05:29, 12 August 2023 (UTC)
IIRC it was set at today minus 8 days (instead of today minus 7 days) to ensure that all AFDs listed were those actually eligible for closure. We did have a big problem with people closing non-snowy XfDs early, assuming that the 7 days was counted from midnight. I see that this has happebned today - for example Wikipedia:Articles for deletion/Alternative versions of Robin was raised at 02:56, 5 August 2023 (UTC) and was closed at 02:34, 12 August 2023 (UTC) - that is, 6 days 23 hours 38 minutes later. Whilst 22 minutes can be seen as a technical quibble, it did not run the full 7 days. --Redrose64 🌹 (talk) 16:50, 12 August 2023 (UTC)
That makes sense, but I think -7 days is more useful; the vast majority of AfDs at -8 will already be closed. Maybe a reminder of the 7 day duration on the AfDs closing today page would be warranted? Actualcpscmscrutinize, talk17:08, 12 August 2023 (UTC)
I remember when AFDs were only open for 5 days, and we had disputes over people closing them early back then, too.
Expecting people to pay attention to tiny little details, like the exact number of hours/minutes the AFD has been open, is unrealistic. If you want them to not close discussions early, you have to make it difficult for them to make the mistake.
I suggest that you consider whether rushing to close an AFD at the earliest possible moment is something we should be doing. We know that it's actively harmful to close a discussion at half a day early, because that causes disputes between editors. The dispute is unpleasant, and the discouragement to AFD closers is damaging. Is it actually harmful to close a discussion half a day after the first possible minute? Who gets hurt in such a situation, and how? WhatamIdoing (talk) 16:40, 14 August 2023 (UTC)
The primary tool for closing discussions, used by the vast majority of closers, is WP:XFDC. This does in fact prompt closers if the discussion hasn‘t been open for the full seven days, so it‘s very hard for me to imagine how it could happen accidentally.
I don‘t think we should be „rushing to close“ as soon as it‘s been exactly seven days. But the script is more useful if it links to a list of AfDs that are actually „closing today“, i.e. where the seven day period has elapsed or will elapse within a few hours. Actualcpscmscrutinize, talk16:50, 14 August 2023 (UTC)
What is more useful about it?
I can understand someone being irritated that the name says "closing today", if they thought that it meant "these AFDs are close later today, so you still have a chance to participate" instead of "these are the AFDs that admins are currently/already in the process of closing today", but that sounds like a reason to change the page title, not to change the contents of the page. WhatamIdoing (talk) 17:16, 14 August 2023 (UTC)
But both apply to the current version. The primary use of this page is, as far as I‘m aware, for closers to see discussions that have recently reached a duration of 7 days. If you go to the deletion log for 8 days ago, most of the time, the vast majority of discussions will already be closed or relisted. I really don‘t see the problem with the 7 day version; if I‘m missing something, please do point it out. Actualcpscmscrutinize, talk18:47, 14 August 2023 (UTC)
It sounds like the 7-day version isn't precisely 7 days 0:00 hours through 7 days 23:59 hours. It sounds like the 7-day version includes discussions that are (at the time they first appear on that page) less than seven days old.
"The vast majority of discussions get closed or relisted as soon as they are eligible" is not what I would call a problem. Are you trying to optimize this for the sake of users who want to close more AFDs personally and are feeling disappointed that there are (apparently) plenty of other people already doing this work? WhatamIdoing (talk) 03:18, 16 August 2023 (UTC)
Neither the previous version nor this one are „precisely 7 days 0:00 hours through 7 days 23:59 hours“, the page just transcludes the deletion log for a specific calendar day. I’m not trying to optimize this for users who are disappointed they can’t close more AfDs, I’m trying to optimize it for users who close AfDs in general, because they’re the ones who use this. Actualcpscmscrutinize, talk07:15, 16 August 2023 (UTC)
Redirects to dynamically-calculated pages don't work, so that would require a bot to edit it every day, which is more effort than value. Perhaps tomorrow there will be fewer AfDs so the page will work again. * Pppery *it has begun...18:29, 17 August 2023 (UTC)
Redirects to dynamically-calculated pages don't work: Well, that's a disappointing limitation. I presumed that if it can be transcluded, it can be targeted with a redirect. Thanks for pointing this out. Actualcpscmscrutinize, talk18:37, 17 August 2023 (UTC)
The Metrolyrics link isn't unfit, it's dead. Unfit is used when the original source has been taken over by a inappropriate site. There's very few uses where it's appropriate, the common one I see that is appropriate is atimes.com that used to be a news sources but now sells steriods. -- LCU ActivelyDisinterested∆transmissions∆ °co-ords° 11:16, 17 August 2023 (UTC)
That's how it looks in Hebrew Wiki (it is in Swedish but never mind). press on the image to see it on full scale
Hello all.
For approximately 16 years (ever since 2007) Hebrew Wikipedia uses a Gadget calls "my links" (look here). It is quite a simple method which let every user to save their own custom side bar links (on the left side of the screen) and makes life much easier.
I ask you, if you approve, to translate this super-easy "pseudo-English" code into WP, as so it would be available for us.
The Gadget adds user custom links to a new section above the Community Section (on the left side of the screen).
how to use it
After one approves “adding user custom links” in the Gadget section inside the Settings, one needs to create “User:<user name> /My links” (i.e. User:Anderssøn79/My links). In order to add a link, one needs to add a new line starting with a “dot/star” (*) and then write down the ref (just as regular [[Page|Alias]] or [URL Alias]).
@Anderssøn79 we generally document our gadgets in the project namespace here. The mediawiki talk:x.js pages are primarily redirected to the project talk, or only really used for edit requests. — xaosfluxTalk15:57, 18 August 2023 (UTC)
After reading a comment about French Wikipedia's style rules for discussion threads, I created my own user stylesheet to experiment with styling discussion threads. It has CSS style rules to define alternating background colours for each nested reply (up to 18 levels), and to add a vertical line to the left of each list (that is, corresponding to one level of nesting).
If you use a recent version of Safari or a Chromium-based browser such as Chrome or Edge, and are interested in trying it out, see User:Isaacl/style/discussion-threads. It's subject to change as I try different things. It works for any page where the reply tool or topic subscription is enabled. (The stylesheet checks for an HTML attribute that the discussion tool features add to mark comments.) A caveat: the styling is awkward with comments made using bulleted list items. All the same, so far I find it helpful to track threads. isaacl (talk) 01:08, 15 August 2023 (UTC)
Is there some tool which can find all instances of dates in an article and convert them all to a uniform format? For example 2023-08-18 > August 18, 2023? RoySmith(talk)18:17, 18 August 2023 (UTC)
An excellent tool, but beware that it likes to insert {{Use mdy dates}} above {{short description}} so they need to be swapped manually. Certes (talk) 19:51, 18 August 2023 (UTC)
I'm confused by this note in the docs: "Functions converting dates within citation templates has been withdrawn now that WM software engine renders date formats automatically". Fixing the dates in citation templates is exactly what I need this for. So what am I doing wrong that the "WM software engine" isn't rendering things correctly for me? This came up in Talk:Bronx County Bird Club/GA1, and I just went and fixed all the dates manually. Was that not the right thing to do? RoySmith(talk)19:56, 18 August 2023 (UTC)
WM software engine no such thing. The citation templates will format publication, archive, and access dates according to the presence of {{use dmy dates}} or {{use mdy dates}} neither of which were in Bronx County Bird Club when I looked. That functionality in the cs1|2 templates is there so that no one has to fuss with date style in the template wiki text. See the template documentation for the {{use xxx dates}} for more detail.
When I've used dates.js, it's formatted all dates, including those within citation templates. I don't recall what combination of existing "use...dates" tags and CS1/2 formats those articles used. I even had to undo the formatting on a couple of dates within URLs to avoid breaking them. Certes (talk) 20:09, 18 August 2023 (UTC)
Oh, my, so I've really made a hash of this then? What I really wanted to do was leave the dates in yyyy-mm-dd format (as the citation tool generated them) in the citation and add {{use dmy dates}} to the top to get them to render the way I wanted them? RoySmith(talk)20:16, 18 August 2023 (UTC)
RoySmith, that is usually the easiest way to fix the rendering a bunch of dates within citation templates. It works only when editing the entire page. Definitely proofread the output for outliers and errors. – Jonesey95 (talk) 20:38, 18 August 2023 (UTC)
Option to exclude drafts and templates from tracking categories?
I've been fixing citations in articles listed in Category:CS1 maint: numeric names: authors list. It would be nice to be able to exclude drafts and templates from that list. I won't go fixing citation issues in drafts, and it seems that no one else is doing that, either, as for many initial letters there are only drafts left. Templates have numerical parameters on purpose, so they don't belong in that list, either. It would be both more satisfying and easier to find articles in need of fixing if the drafts and templates could be excluded, either from the category entirely or from my view of its article list. Is that possible? Joriki (talk) 12:44, 16 August 2023 (UTC)
Oh, I thought MediaWiki messages with category names like MediaWiki:Cite-tracking-category-cite-error had to be pure text but it allows wikitext. It's undocumented whether a message allows wikitext. That makes it a change local administrators can easily make. We could make subcategories for drafts, templates, project pages, a miscellaneous category, or whatever we want. I doubt sortkeys would be possible but I don't know how MediaWiki processes the message. It's possible that name|sortkey or name{{!}}sortkey would work. I see you have a suggestion at Category talk:Pages with reference errors#Draft, Wikipedia, Template, and Help. If you really want sortkeys to group other namespaces at the end like [21] instead of subcategories then I could make a test at testwiki: without interrupting Wikipedia. PrimeHunter (talk) 13:04, 17 August 2023 (UTC)
How to avoid collapse template getting indented on t/p
During discussion pages usually I reply by clicking on reply link since that automatically selects requisite indenting. Problem comes when I use Collapse template {{collapse top|Reply with details 1}} .. {{collapse bottom}} in my own comment to keep discussion concise and easy navigable for other users. Indenting is automatically applied before collapse template too and there after subsequent text below {{collapse bottom}} also gets hidden / invisible until remove that extra indenting manually with another edit.
Many times my internet speed slows down or Wikipedia gives error of being busy for maintenance and it takes some time before I succeed in removing such unrequited indenting. (And I am experiencing internet speed problem just now too) And I am always worried some later users may misunderstand me.
Is there any way to update that collapse template so as later text will not become invisible even if indenting is there or any way around for me to avoid such difficult situation?
Where can I write the original pronunciation of a surname?
I noticed some pages (Lionel Messi, Javier Mascherano) do not sport the original Italian pronunciation of the persons' surnames anymore. Is there a guideline about where to write the original surname pronunciation? On Wikidata perhaps? In Elias Lönnrot I wrote it in a footnote.-- Carnby (talk) 19:33, 18 August 2023 (UTC)
@Xaosflux Honestly, I can't find anything about non-English names originating in a different country, such as Messi, Mascherano, Di María, Bergoglio, Milei (Argentine people of Italian descent), Antetokounmpo (Greek person of Yoruba descent), Lönnrot (Finnish person of Swedish descent)...-- Carnby (talk) 21:07, 18 August 2023 (UTC)
That's not a technical question. I tend to find it unnecessary unless the subject has strong ties to the "original" language (they're fluent in it, they're a first-gen immigrant, etc.). Nardog (talk) 01:43, 19 August 2023 (UTC)
Something keeps leaking memory. It might be one of the scripts I'm using, but I'm not sure. Wikipedia tabs (to the exclusion of all others) inevitably glitch out when left sitting for a while and start using upwards of 1.5G of RAM. How can I find out what exactly is causing this? I'm on Vivaldi 6.1.3035.257, but this has been going on for months. 〜 Festucalex • talk09:52, 18 August 2023 (UTC)
Unless there's an endless loop in a user script or something, a script itself shouldn't cause a memory leak, in my opinion. It's my understanding that memory leaks are usually a problem with the browser's code. –Novem Linguae (talk) 01:03, 19 August 2023 (UTC)
Although they're not the same as a traditional memory leak, store gobblers can still occur with Javascript. (I do not know, though, if this is the case in this situation.) For example, a script could continually register new event listeners, thus consuming memory and hampering performance. isaacl (talk) 01:45, 19 August 2023 (UTC)
Not a webdev, but my understanding is that this became untenable after browser makers and web devs came to see the web as a "platform" which be as powerful as, and eventually supplant, a full desktop OS's native runtime environment (e.g. wasm, and recently WebGPU). When Apple changed Safari to auto-crash tabs that used too much memory, people complained that it broke mission-critical sites like Zoom, and even mundane stuff like Disney+, so Apple relented. DFlhb (talk) 10:08, 19 August 2023 (UTC)
Not sure what you are referring to by "this"? Memory leaks for long-running programs have always been seen as a problem to be solved (although the rate of increase will affect how the work gets prioritized). Store gobblers from a Javascript file served by a website can't be fixed by the browser vendor, so the browser can only balance protecting other tabs from being affected by interfering with the resources allocated to the problem-causing tab. isaacl (talk) 13:23, 19 August 2023 (UTC)
"This" = website memory leaks being automatically prevented by browsers. Browsers can't do that without preventing high resource use for legitimate, non-bug purposes. DFlhb (talk) 14:10, 19 August 2023 (UTC)
I am using Firefox on Windows 10 Pro PC. Latest standard versions on both. This does not happen to me on the Edge browser. Whether I am logged in or not. I logged out of Wikipedia and I still have the problem with Firefox. So it is not my JS and CSS files. I also disabled all my Firefox addons and then closed all my Firefox windows. I launched Firefox and I still have the problem with Firefox. No addons and not logged in.
The problem tables all use templates within the table. But it is not a problem with tables using {{flaglist}}.
Open visual editor. Click to the right of the either of the 2 tables. Note the bizarre result of the row moving over one cell. Click to the right of a different row and note that row moving over one cell. If you click above or below the table the problem goes away.
I opened visual editor for Help:Table and tried clicking to the right of many tables. Closer and farther away. I have noted the problem there with only one table so far. Have to go there to use visual editor on this table:
I see the problem only with the middle row there. Click to the right of it, and click closer until the row moves over one cell. It doesn't happen all the time.
I removed the templates from a few rows of one of the tables in the list article. This solved the problem. And the small table above only has the bug in the middle row using templates. --Timeshifter (talk) 15:00, 14 August 2023 (UTC)
That search query has some false positives. The first entry I found with #:~: is here, but that one seems fine. It doesn't seem to identify the user. Got any other examples? Are you requesting help to clean these up? –Novem Linguae (talk) 01:15, 19 August 2023 (UTC)
The purpose of this is not to identify users (although I suppose that may be possible?) - rather, it's to link to a specific piece of text on a page. (That said, it is one of those things that Google kind of unilaterally decided to add to Chrome, with varying lack of support across other browsers.) LittlePuppers (talk) 01:30, 19 August 2023 (UTC)
Template:Citation Style documentation#url says to omit tracking parameters but I wouldn't call this a tracking parameter, and it also says "the link may point to the specific page(s) referenced". I think that mainly refers to PDF files where it's possible to give a page number in the link. Do we have a guideline saying to omit Google's text= in cases where supported browsers will jump directly to the part of the page being used in a reference? I guess nearly all cases where text= was left is just users who clicked a Google link and copied the full url without knowing the meaning, but a guideline could lead to more cleanup where it's removed. PrimeHunter (talk) 08:59, 19 August 2023 (UTC)
The octothorpe isn't a special regex character so I'm not sure how that works, but then the search regex works strangely. — Qwerfjkltalk21:00, 20 August 2023 (UTC)
Not a Wikipedia issue, exactly, but I suppose people here would know the answer.
meta:Special:CentralAuth/Nyttend gives my account registration date as 30 March 2008, apparently because that's when I linked my 2-year-old en:wp, commons, and de:wp accounts. For us old-timers, who remember when you had to register accounts separately at the different projects, is there anywhere that provides each account's creation date? Or if I'm curious, do I have to look at each wiki's log page? Nyttend (talk) 11:37, 21 August 2023 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
The new version of MediaWiki will be on test wikis and MediaWiki.org from 22 August. It will be on non-Wikipedia wikis and some Wikipedias from 23 August. It will be on all wikis from 24 August (calendar).
Is there any technical reason that we don't have a namespace devoted to scripts? It occurred to me that we have spaces which encourage collaboration among editors for coding Modules, and Templates, and we even have Draft space for collaborating on articles-to-be, we don't have a space devoted to scripts, and so they are developed in User space, typically as a subpage. A negative consequence of this, imho, is that for better or worse, editors are "respectful" (not quite the right word) of user pages "belonging" to other users, and might shy from editing other users' pages more than they might a Draft, a Module, or a Template. I've been around long enough to know that no page belongs to anybody, but it's kind of an emotional thing, almost; I'm much less likely to jump onto your User subpage and start improving "your" draft, "your" template, or "your" module. But in all of those cases, the user has the option to move them to Draft:, or to Template:, or to Module: space. Not so with scripts, and I fear that this has negative consequences for collaboration on scripts. Maybe I'm all wet, and if script writers out there tell me that they happily jump in to scripts regardless what user subpage they happen to live on, then I'll be happy to hear it, and we can close this. But it almost feels like people have to ask permission or something before messing with a user script.
There might be some statistics we could look at to try to assess whether this is, or isn't a problem, and I've asked the nice folks over at WP:QUERY to see if they can enlighten us, and will link this discussion and report back. Thanks, Mathglot (talk) 02:04, 21 August 2023 (UTC)
For me, it's not about being respectful by choice. When I go to User:Mathglot/ShowRevisionID.js, I see "View source" instead of "Edit". I can't edit that page, even if I am 100% confident in my proposed change. I have Template Editor rights but I am not an admin. If only admins can edit other editors' .js pages, it seems highly unlikely that a collaborative culture of script editing would develop in User space. If that makes sense. – Jonesey95 (talk) 02:15, 21 August 2023 (UTC)
Admins cannot edit .js pages, only the owner or interface admins. That's a security issue because js runs on the user's browser and is an uncontrollable language able to do a lot of damage. By comparison, templates and modules use a well-defined and locked down language that runs on monitored servers. Johnuniq (talk) 02:22, 21 August 2023 (UTC)
Yes, it reinforces your point but it also reinforces mine, namely that js pages are a security problem and opening them up to more than a dozen or so people will not and should not happen. Johnuniq (talk) 02:37, 21 August 2023 (UTC)
Whoa, I'm surprised! That would seem to argue strongly in favor of a new namespace, and a new privilege. Why shouldn't trusted editors be able to collaborate on scripts as they do on modules and templates? I don't really see that the security issue is a problem, because we all get to import or otherwise install somebody else's script on our common.js, and the same thing would apply, even if we imported it out of Script space, instead of user space. The only difference would be that we'd have more trusted eyeballs and hands on the scripts in script space; that seems like a win-win. (edit conflict)
Putting it another way: what's better: limiting controlling potential damage to one editor-owner (plus 12 IA's), or allowing a coterie of trusted js editors to monitor and look over all the js files, regardless where they live? And additional protection could always be applied to any given page, as is the case now. Mathglot (talk) 02:48, 21 August 2023 (UTC)
A permission model allowing user-defined groups, for example, doesn't need a new namespace. Without some ability to set editing restrictions, the security issue is indeed a significant problem: I don't want anyone changing my personal user scripts without my authorization as that would allow them to trigger the execution of arbitrary Javascript code in my browser with my credentials. isaacl (talk) 03:01, 21 August 2023 (UTC)
This collaberative space could live in the MediaWiki namespace. An script living there does not make it a gadget, it is both it's location there and that it is specified in MediaWiki:Gadgets-definition that does. An distinction between an collaberative script and an gadget could be made with an prefix, Gadgets allready have one, or at least the first script in a row (like with MediaWiki:Gadget-refToolbar.js and MediaWiki:RefToolbar.js).
It is really unlikely that an new user-right would happen. Interface admins where added by the WMF and they are not interested in slacking the requirements on it. 2FA is a known deterrant for people to get IA rights and will continue being so.
It would be nice to have an database query on scripts by inactive users (that metric is well defined nowadays), so that those scripts could be moved from Inactive_user/script.js to new_maintainer/script.js, where a new maintainer is interested. Snævar (talk) 11:15, 21 August 2023 (UTC)
There are use cases where user-defined access groups are helpful to allow greater flexibility in managing who can update a specific page, though possibly more so with non-WMF deployments of MediaWiki. (For an example of how this works with a different Wiki platform, see Foswiki's documentation on access control.) isaacl (talk) 16:43, 21 August 2023 (UTC)
Isaacl, Re: "...as that would allow them to trigger the execution of arbitrary Javascript code in my browser": Don't we already have that situation now, in the case of imported scripts in common.js? Sure, it's "my" js, but I'm importing a bunch of useful scripts, and I'm trusting that the changes they make to the scripts behind my back are beneficial. I import them because I trust them, even with changes I don't see, but even a well-intentioned script change might be destructive. How is it better to have that under the control of one person, rather than a community of trusted script editors, who might jump in from any time zone when they noticed something awry? Seems to me the supposed security of having only 12 IA's isn't really there, it's a mirage, unless you don't import any scripts at all, and then you're safe.
But, your objection is still a fair one, so let's see what can be done. What if imported scripts were disabled automatically every time they were modified, and then either you could reënable it yourself (just for you, redeclaring your trust, either after looking at the script, or just trusting its author), or it would be globally reënabled maybe after an Rfc-like script Talk page discussion where the script author would explain/defend their change, and the community would respond approve, decline, needs-work? I know that bots have to go through an initial approval and beta-test period, before being allowed to run. What about afterward: is initial approval carte blanche to do whatever modification you want to the code as long as you don't call it a "new task"? Who decides what kind of bot-code change is too big to allow without approval? How is importing a script which runs in your browser and is only editable by one user, safer than one that is editable by a community of trusted users?
If we had a community of privileged script writers and automatic disabling of scripts every time they were changed, along with some mechanism to reënable them, would that satisfy your objection regarding scripts written by others running in your browser with your credentials? If we're halfway-there, and there's just a missing piece, what is it? Mathglot (talk) 01:18, 22 August 2023 (UTC)
Currently, I control when I modify my common.js file, and make an explicit decision to trust one specific user. Any change to allow users to modify other people's user scripts will be different than the current situation. I think having user-defined access groups would provide a framework for defining circles of trust, both regarding who can alter a user script, and what circle someone needs to trust before making use of the script. isaacl (talk) 02:01, 22 August 2023 (UTC)
You can already import a specific revision of a script if you don't want to risk the author going rogue or being compromised, although you then have to review each update or miss bugfixes and improvements. Everything about user scripts is a tradeoff between convenience/innovation and security. The fact we let users execute arbitrary code and that code is out in the open is unthinkable to most engineers. Reviewing each script and each change would be a tremendous undertaking neither WMF nor the community can dream to afford. But they make your life on wiki a lot easier, don't they? Nardog (talk) 02:05, 22 August 2023 (UTC)
Extra space in between namespace and talk page name
In phab:T315893, for folks that have enabled Special:Preferences -> Editing -> Show discussion activity, the editing team has decided to add a space after the colon to every talk page title. It sounds like they have plans to roll this out to non-talk pages and possibly for non-"Special:Preferences -> Editing -> Show discussion activity" users eventually. An example of this space can be seen by clicking here. The editing team has stated that this needs to be overridden on a per-wiki basis if we don't like it, I assume using MediaWiki:Common.css.
Thoughts on this change? Should we consider overriding it? I'd like to override it, personally. As a programmer it bugs me that the software is suggesting a title that isn't the "correct" title. This isn't a change we asked for. It seems confusing. Happy to hear other thoughts though. If there is consensus to change it in this discussion, I will put in an {{IAER}} to MediaWiki talk:Common.css after a couple days. Thanks. –Novem Linguae (talk) 12:44, 22 August 2023 (UTC)
It's not a space character. It's a CSS-generated gap. Copying and pasting works fine. There was a bunch of hand-wringing by a couple of editors on the phab ticket, but I haven't noticed anyone freaking out about it here. It's been deployed for a while. The weird part about it is its inconsistency: it is present in Talk pages but not here at VPT, even though the function of the colon separator is the same in both places. It should either be present everywhere or nowhere. – Jonesey95 (talk) 14:01, 22 August 2023 (UTC)
The inconsistency is exactly why we should hide it. If you were serious about making our NS-prefixed page names conform to orthography, you wouldn't do it in just a few namespaces or tweak just the heading. You would consult experts like the language committee, assess the breath and risks of the change, allocate resources, and publicize it to the communities. The fact they aren't doing that says this is an ill-conceived idea. Nardog (talk) 15:50, 22 August 2023 (UTC)
Or you could just add the CSS spacing to all namespaces and see what happens. It does not seem to have caused any harm in its beta deployment. I am frustrated by a lot of beta stuff that the WMF deploys and then leaves 80% done (cf Visual Editor, Vector 2022), but this seems like a small thing. If it causes trouble, we can remove the spacing locally, it appears. – Jonesey95 (talk) 16:01, 22 August 2023 (UTC)
Add .mw-page-title-separator{margin-right:0.25em;} to your CSS and see how long before you have urge to write page names with the extra space or to see the rest of the interface the same way. Nardog (talk) 16:27, 22 August 2023 (UTC)
Many other editors and I have seen the spacing on talk pages for a while now. I have no such urge and have seen no evidence of such an urge. Claims that the sky is falling require evidence. – Jonesey95 (talk) 17:38, 22 August 2023 (UTC)
Agreed with @Jonesey95 that the consistency should be fixed, but otherwise it doesn't bother me. The whole concept of namespaces is one of the most confusing things about MediaWiki. This sort of helps grasp the concept because it clearly differentiates the namespace from the title. Yes, the visual spacing is jarring if you already understand namespaces, as you wouldn't type it out like that, but as it happens, it doesn't actually matter! Talk: Polar Bear still works :) That said, I do share the same concerns with @Nardog at phab:T315893#9059239 that the sake of consistency might over time might lead editors to always use spaces after colons. Before you know it, we've got yet another section added to WP:MOS, followed by tons of WP:AWB edits to make it all consistent. I don't think any of us want that, but I also think it's too early on to really say how problematic it will be. Thus far it seems to have caused few to no problems, so I'm led to believe it's one of those changes that we'll all simply get used to over time. The extra spacing bothered me when it was first deployed, but I quickly got over it, and now I even sort of like it. It's just easier on the eyes.
At any rate, we can very easily override it with sitewide CSS if needed, so I'm not really worried. I say let them upstream it to MediaWiki Core if they want – and by doing so fixing the consistency issue Jonesey95 mentioned – then we'll see how folks react. Once it effects all namespaces and all MediaWiki installations, we'll surely know if it's a problem or not. I'm aware of the issues with orthography in other languages, but here we're only debating it for English, where it doesn't seem to be that big of an issue. — MusikAnimaltalk15:54, 22 August 2023 (UTC)
I submit that "tons of WP:AWB edits to make it all consistent" are exactly what's going to happen, and what if the community says no and we have to revert the edits again? Or worse, a drawn-out RfC? I certainly don't want to see editor hours wasted like that. NL's proposed edit to the site-wide CSS is a great way to be proactive and prevent it. Nardog (talk) 16:25, 22 August 2023 (UTC)
Will the gallery tag support the upright parameter?
Currently the <gallery> tag and Template:gallery which implements it, can be sized using pixel values in the |height= or |heights= parameters. The help page says that Images displayed by the <Gallery>...</Gallery> tag do not obey user viewing preferences. Is there any plan to support user preferences and the |upright= parameter as image thumbnails currently do? Or is there a reason why this will not happen? Rjjiii (talk) 03:17, 22 August 2023 (UTC)
No one has touched its code in years. The last thing that as added to it was the slideshow mode, but its core fundamentals are ancient. As such it has not kept up with other parts of the code. —TheDJ (talk • contribs) 11:25, 22 August 2023 (UTC)
Well dang. I was thinking that galleries might be a cleaner way to juxtapose images, compared to Template:Multiple image or a single image created for comparison. Thanks for all of the explanation though. Rjjiii (talk) 02:43, 23 August 2023 (UTC)
How to add Infobox constituency quickly?
Hi Techie, Check this link i want to add Infobox with details like name, cons No, Map, electors, current MLA etc., i have some knowledge about auto wiki browser. I cant use easily this tool. Kindly suggest any alternate tool/method available. Thanks in advance IJohnKennady (talk) 18:58, 23 August 2023 (UTC)
@Edward-Woodrow: MDN statesNote: You can use href="#top" or the empty fragment (href="#") to link to the top of the current page, as defined in the HTML specification. What browser are you using? It works fine for me in Firefox.
I see a link to #footer, which works for me; it appears that something (probably MediaWiki) adds <footer id="footer" class="mw-footer" role="contentinfo" > at the bottom of the page. What skin are you using? LittlePuppers (talk) 21:42, 23 August 2023 (UTC)
Okay, so I see links to the bottom using #footer, #footer-info, and #skip-to-bottom-anchor. Those appear to be in all skins, Vector (at least the tolerable version, which I also use), and {{Skip to top and bottom}}, respectively. I'd change it to #footer. LittlePuppers (talk) 21:49, 23 August 2023 (UTC)
To make the header independent of the context in which it is transcluded, it can also position a <div> element at the bottom of the page with a specific ID and use it. But either way, the message ought to only be shown when there are questions to be read and skipping to the bottom makes sense. isaacl (talk) 22:31, 23 August 2023 (UTC)
The same way the {{skip to top and bottom}} template does it: use CSS style properties to position it. Yes, the Teahouse page already inserts a <div> with a specific ID that can be used, but this makes the header dependent on the Teahouse page always transcluding {{skip to top and bottom}}. I think, though, controlling the message's display is more important than making the header manage its own target ID. isaacl (talk) 22:46, 23 August 2023 (UTC)
Searching for exotic whitespace
I'd like to search for articles containing strings like "=<whitespace>http" .. where "<whitespace>" are characters listed at whitespace character (except for the normal space). For example to find "thin space", what would the search string be? I tried insource:/[=](\u2009){1,}http/ .. because the codepoint is U+2009 and according to this tutorial it might work, but does not. CirrusSearch docs has some info but nothing I can see helps. -- GreenC01:10, 24 August 2023 (UTC)
\u... isn't going to work. Googling "regular expression" my problem is of very limited help for Elasticsearch (which is what CirrusSearch uses as its backend); you're going to get results for the PCRE variety of regular expressions that are the de-facto standard, and the flavor Elasticsearch implements is much more limited and uses somewhat different syntax. (Upstreamdocumentation.) Searching for the literal character, like this, gets me correct-looking results. Or, at least there was a literal thinspace in the source of the articles I spot-checked. (I was surprised that there isn't in either whitespace character nor thin space). —Cryptic02:06, 24 August 2023 (UTC)
OOUI is not only a big library but one WMF is moving away from. Doesn't strike me as something worth making every reader of some of the most visited articles on the wiki download just so the radio buttons look slightly prettier. Nardog (talk) 01:48, 22 August 2023 (UTC)
Nardog, WMF moving away from OOUI? I was unaware, where can I read more about that? What is the replacement? — Alexis Jazz (talk or ping me) 09:57, 22 August 2023 (UTC)
TheDJ, I wish someone had told me this before I invested too much time figuring out OOUI. And now Codex. It just doesn't seem to work. "Uncaught ReferenceError: require is not defined" is all it can throw and none of these code examples has any context on how to actually use any of it. :-/ I just don't get it. The only example I could find here is User:Jdlrobson/rl.js but I don't understand what that code is doing. — Alexis Jazz (talk or ping me) 17:44, 22 August 2023 (UTC)
@Alexis Jazz I doubt OOUI will ever be removed, and I'm certain it won't happen in the foreseeable future (not until someone rewrites VisualEditor and a bunch of other things), so if you've grown to like OOUI, you can keep using it. (But if you always hated it, now you're vindicated!)
I'm not sure what's the context for your require question, but you can get that function from mw.loader.using like so: [32]
(or you can use mw.loader.require instead, but that is "publicly exposed for debugging purposes only and must not be used in production code" [33]).
Codex is more complicated than OOUI, unless you've already used Vue.js, in which case it's more intuitive. It's also quite annoying to use from gadgets, since ideally it depends on a bunch of server-side processing. I haven't tried doing that myself yet, and I hope someone will document how to do it at some point (or make it easier).
If you're just trying to get things done, in a way that will be stable for years, I recommend the "CSS-only version" of Codex (e.g. [34]), which is way easier to understand than everything else, as long as you don't mind writing out a bunch of CSS classes in your JS code. Matma Rextalk20:13, 22 August 2023 (UTC)
Matma Rex, I just wanted to create a "hello world" dialog. I've added your line so it no longer errors, but I still get no dialog: Wikipedia:Sandbox (revision 1171734233) Even if that worked, it's a ludicrous amount of code to produce a "hello world" popup. As for OOUI, I have a kind of love-hate relationship with it. In some ways it's too limiting and I end up applying CSS hacks which can end up harmful, and some stuff is just really confusing. (IIRC reading/writing the options from comboboxes was confusing but it's been some time) The icons OOUI provides I don't use at all, they add considerable loading time and have a more restrictive license than my scripts, so I just designed my own icons. But these proved more difficult to apply to OOUI elements. From what I read, with Codex you can load only the icons you actually need which would seem nice. Some parts of OOUI (long names of stuff) add considerably to the size of scripts. But in general I like OOUI. I don't have to think as much about various skins and styling. And once you get the hang of it it's okay to work with. — Alexis Jazz (talk or ping me) 23:33, 22 August 2023 (UTC)
@Alexis Jazz I got your example working with three changes: [35] (just for educational purposes)
You don't need to create a HTML <template> element. The example code uses that, because the server-side processing I mentioned detects this syntax and puts it elsewhere. If you don't have that (like in a gadget), you need to use a normal string, and pass it to the component.
You don't need to (and can't) do anything with module.exports. That's used when you have a ResourceLoader module consisting of multiple files. mw:ResourceLoader/Package files
You were defining a reusable component, but you were not using it anywhere. In theory, once you had your DialogBasic component, you could use it as <dialog-basic> inside other components, same as you're using CdxButton / <cdx-button> (I'm not sure if that requires some other boilerplate) – e.g., in the context of MediaWiki, imagine a reusable namespace selector dropdown defined that way, or something. But if you want to display it, you need to do it yourself. Luckily this is just one line.
Matma Rex, I confirm it works, thanks! As also stated by egardner on phab:T344680 it's not really streamlined enough ATM to recommend for use in gadgets or userscripts. However, it's nice to at least know how to make it work, even if it's not particularly practical ATM. — Alexis Jazz (talk or ping me) 02:19, 24 August 2023 (UTC)
@Alexis Jazz: Thank you very much for looking into this. I'm unfamiliar with Mediawiki software so cannot comment on what the correct approach is but am happy to beta-test it. https://commons.wikimedia.beta.wmflabs.org/wiki/Template:Switcher has adequate space between each radio button and its label, though there is much larger vertical separation between rows than on Template:Switcher/doc#Example: Firefox's Inspector shows that the wmflabs version has each label in a separate div where the original template has just label elements. Cheers, cmɢʟee⎆τaʟκ16:05, 22 August 2023 (UTC)
Math parsing error
Hi, At article D-arabinitol 2-dehydrogenase, all I did was remove the Orphan tag at the article top, and now it shows in red:
Failed to parse (SVG (MathML can be enabled via browser plugin): Invalid response ("Math extension cannot connect to Restbase.") from server "http://localhost:6011/en.wikipedia.org/v1/":): \rightleftharpoon</ref>
For this specific case, math seems like overkill for just a single symbol that is available in plain-text (or the {{eqm}} char template if you prefer). DMacks (talk) 17:23, 22 August 2023 (UTC)
In the Sentinel program article there are a couple of quote blocks, using cquote. The second of these has an odd character at the start. This character does not seem to be in the source text. Look for "Which weapons", I see a rectangle in front of the "W". Does anyone else? Maury Markowitz (talk) 16:54, 24 August 2023 (UTC)
I have removed a stray control character.[36] It's browser-dependant what, if anything, is shown. You can copy-paste something to the "Characters" field at https://r12a.github.io/app-conversion/ (not Wikimedia affiliated) and click "View in UniView" if you want info about special characters. PrimeHunter (talk) 17:23, 24 August 2023 (UTC)
Hyphens gadget
I'd like to have a gadget to break words with hyphens. This feature is available in CSS and it looks nice with paragraph text justification turned on. The CSS code for the gadget would look as following:
You should probably ask tech questions about Commons on Commons. They have enough savvy users there to answer. Izno (talk) 02:12, 26 August 2023 (UTC)
Is importScript asynchronous?
I use AutoEd with a custom set of functions in my common.js. Rather frequently AutoEd fails to run, with a warning left in my browser console along the lines of jQuery.Deferred exception: autoEdISBN is not defined ReferenceError: autoEdISBN is not defined. My suspicion is that importScript('Wikipedia:AutoEd/isbn.js') is importing the autoEdISBN function asynchronously, which creates a race condition where AutoEd might try to invoke it before it has finished importing. Is that right? And if so, is there some way to wait until importScript is finished? rblv (talk) 01:51, 26 August 2023 (UTC)
No. It was not made async until Krinkle's adjustments last year to "undeprecate" it. This was one of the reasons it was not made a shim for mw.loader.load originally. Izno (talk) 15:42, 26 August 2023 (UTC)
That is not true, or rather – it is a myth. ImportScript was always async. It shared mostly the same code as the mw.loader family – the approach being to insert a <script> element into the page, which causes deferred loads in all browsers. But unfortunately a few users took to describing importScript as synchronous and "bad" without looking into its source code. – SD0001 (talk) 16:39, 26 August 2023 (UTC)
Thank you, mw.loader.getScript looks like what I need. Unfortunately I think there is an inherent race condition in how AutoEd handles user-defined plugins. I attempted to fix User:Rublov/common.js by using getScript to asynchronously load the AutoEd plugins, and then attaching a callback to the Promise object inside of autoEdFunctions to ensure that the code doesn't run until the scripts are loaded. AutoEd however treats autoEdFunctions as synchronous and reloads the page as soon as it returns, which is often (always?) before the async callback in autoEdFunctions has had a chance to run.
Unfortunately I think there is an inherent race condition in how AutoEd handles user-defined plugins. AutoEd is ancient and I'm surprised it's maintained much less functional. I would not be surprised if there was one. Izno (talk) 21:31, 26 August 2023 (UTC)
Although Wikipedia:AutoEd/Customization implies the core script should be imported first, the preset files actually defer loading the core script until after the page DOM model is "ready" (however jQuery determines this). Thus I suggest structuring your common.js file to match the preset files, copying how it loads the core script (such as the code from complete.js, line 39 to line 41). I'm not sure if that's still too soon; if so, then instead of loading the core file as in line 40, you can invoke a promise like the one you have in your current common.js file, and in its callback, load the core file. This will result in a clear sequence: document ready → helper files are loaded in some order → core file is loaded. isaacl (talk) 22:06, 26 August 2023 (UTC)
Ah, I had assumed it was necessary for the core script to be loaded first. If not, then it's possible to do everything in the right sequence. Thank you! rblv (talk) 23:15, 26 August 2023 (UTC)
I looked at some of the helper files—they define standalone functions that don't use anything from the core script and don't execute any code, and I saw the preset files loaded them first which confirmed it would work in that order. You're welcome! isaacl (talk) 23:38, 26 August 2023 (UTC)
Dated maintenance categories
Wanted to ask about an issue that I've noticed with dated maintenance categories. This isn't the only time this has happened, and instead there have been a lot of examples of this of late, but I'll raise Category:CS1 maint: DOI inactive as of December 2022 as an example since that's the most recent case I've located.
What happened is that it was deleted as empty on August 1, but then sometime in the next week it somehow became repopulated, with the result that I had to undelete it on August 7 to get it off Special:WantedCategories, but then sometime in the week after that it was emptied again and thus had to be deleted a second time.
Now, obviously, there's no value in tagging dated maintenance categories for the "keep even if empty, because a possibility of future reuse exists" option that general undated maintenance categories sometimes get — but I don't see any great value in Wikipedians going through a looping delete-undelete-redelete cycle on them either. So is there any alternative way that could have avoided this, such as having the maintenance template only generate dated categories if they actually exist, and substituting a generic undated "crapcatcher" category if they don't, so that such situations are prevented from showing up as "wanted" redlinks? Bearcat (talk) 13:03, 20 August 2023 (UTC)
I guess a page reappeared in the category because it was reverted to a revision with that dated tag. A template would have to use #ifexist to omit red categories but it's an expensive parser function with 500 total allowed per page. That could cause problems on pages with many such templates. It would also delay creation of new maintenance categories. I would like a MediaWiki feature like regexes in a MediaWiki message to say that certain category names should be treated as hidden categories if they don't exist. Then normal readers and registered users with default settings wouldn't see an ugly red link in the category list. It wouldn't solve your problem but I think this is a more important problem with red maintenance categories. PrimeHunter (talk) 14:41, 20 August 2023 (UTC)
Umm, are you sure? The only dated maint cats that cs1|2 populates are subcats of Category:CS1 maint: DOI inactive. Presently, there are ~2k articles that have |doi-broken-date= (all of them in Category:CS1 maint: DOI inactive as of August 2023). Since the dated maint cat is dependent on |doi-broken-date=, the cs1|2 module would only have to call the expensive function mw.title.new ('Category:CS1 maint: DOI inactive as of <month> <year>').exists when |doi-broken-date= is found. I doubt that there will be more than 500 of those parameters in a single page. When |doi-broken-date= is found without a matching extant category, the module can emit the red-linked cat plus a blue-linked cat for AnomieBOT which can then somehow create a blue-linked cat from the red-linked cat. Does AnomieBOT add all of the wikitext presently in the category page itself or is that an exercise left to editors?
I thought the bot would recreate a deleted category if it were repopulated? I assume this wouldn't require a lot of work, just a periodic check of "TOPIC as of MONTH YEAR"; there are a lot of possibilities of course (272 months have passed since Wikipedia was set up, and that's for each topic category: 272 "unsourced articles as of", 272 "NPOV issues as of", etc), but it shouldn't be hard for a bot. Nyttend (talk) 11:42, 21 August 2023 (UTC)
I'm beginning to think that an automated solution utilizing AnomieBOT doesn't fit here. The example category, Category:CS1 maint: DOI inactive as of December 2022, is not named properly for AnomieBOT (it wants ... from <month> <YYYY>). It isn't clear to me how to apply either of {{Monthly clean-up category}} or {{Monthly maintenance category}} where the desired category title uses a date that isn't current-month and current-year.
I have to ask, is there any value in the template actually generating dated monthly categories at all? Given that as of right now the dated monthly for August is the only one left, is there a reason why it can't just throw everything into one undated category and stuff anything that would cause problems like this to even occur in the first place? Bearcat (talk) 20:33, 26 August 2023 (UTC)
I'm not doing anything wrt to dated DOI categories. Someone else is though. The solution here really is that when some old broken DOI is found, just run citation bot on the page and it'll update to a modern date (or get fixed). Headbomb {t · c · p · b}00:57, 27 August 2023 (UTC)
Visual editor confusion when combining lang template with a wikilink
I used the Visual editor to add a wikilink to a lang template and vice versa, and it looked fine in the visual editor, but bad in the rendered article. See references to "Du Dingyou (杜定友; 1898–1967)" in Ligature (writing)#Chinese ligatures and Guangzhou Library#History
notice the category wikilinks within the Du Dingyou wikilink which MediaWiki rejects. Here in Wikipedia namespace, what you wrote works because categories are suppressed by the template:
The mobile site has a different layout, narrower content and does not use the full width like the desktop site does.
Here's my problem:
If someone copies a link on mobile and posts it online somewhere, and someone on desktop clicks it, they'll get sent to the mobile site. Even though they are on desktop.
There should be a redirect from the mobile site to the desktop site if the site detects you are using a desktop browser.
Question: Is there a tool for seeing "most-pageviews days" for an article?
We have a great tool for looking at pageviews over a given stretch of time, for example three-year pageviews of Sean Connery, but I am curious as to whether there is a way to see, for example, the top twenty individual days for pageviews for an article. BD2412T01:30, 23 August 2023 (UTC)
I'm not sure of such a tool but, as a workaround, it can be done by using Download > CSV on that page and then sort by the pageviews column in Excel, OpenOffice Calc or other software. Simeon (talk) 13:12, 26 August 2023 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Recent changes
As part of the changes for the better diff handling of paragraph splits, improved detection of splits is being rolled out. Over the last two weeks, we deployed this support to group0 and group1 wikis. This week it will be deployed to group2 wikis. [37]
All Special:Contributions pages now show the user's local edit count and the account's creation date. [38]
Wikisource users can now use the prpbengalicurrency label to denote Bengali currency characters as page numbers inside the <pagelist> tag. [39]
Two preferences have been relocated. The preference "Enable the visual editor" is now shown on the "Editing" tab at all wikis. Previously it was shown on the "Beta features" tab at some wikis. The preference "Use the wikitext mode inside the visual editor, instead of a different wikitext editor" is now also shown on the "Editing" tab at all wikis, instead of the "Beta features" tab. [40][41]
Changes later this week
The new version of MediaWiki will be on test wikis and MediaWiki.org from 29 August. It will be on non-Wikipedia wikis and some Wikipedias from 30 August. It will be on all wikis from 31 August (calendar).
All right-to-left language wikis, plus Korean, Armenian, Ukrainian, Russian, and Bulgarian Wikipedias, will have a link in the sidebar that provides a short URL of that page, using the Wikimedia URL Shortener. This feature will come to more wikis in future weeks. [42]
Future changes
The removal of the DoubleWiki extension is being discussed. This extension currently allows Wikisource users to view articles from multiple language versions side by side when the <=> symbol next to a specific language edition is selected. Comments on this are welcomed at the phabricator task.
A proposal has been made to merge the second hidden-categories list (which appears below the wikitext editing form) with the main list of categories (which is further down the page). More information is available on Phabricator; feedback is welcome!
I'm curious if anyone knows of a discussion or a checklist that led to the decision to declare that the visual editor is no longer in the beta stage of software development. Until one of these recent updates, the Editing tab said something like "Disable the visual editor while it is still in beta". Was there a big set of bug fixes? Is VE still inserting nowiki tags where it should not (that diff is from four weeks ago), a long-standing bug? – Jonesey95 (talk) 15:53, 28 August 2023 (UTC)
The software is stable and has been for the better part of a decade at this point. It should have been "out of beta" long ago. Izno (talk) 17:37, 28 August 2023 (UTC)
Thanks. I agree with the latter part of your statement. I will keep my eyes out for new instances of still-open bugs that I filed half a decade ago. Maybe they have quietly been resolved but not closed in Phabricator. That would be nice. – Jonesey95 (talk) 18:21, 28 August 2023 (UTC)
Tangentially related: the edit count shown is the number of edits actually made by yourself, and excludes transwikid edits. As an example, see de:Spezial:Beiträge/Redrose64 - this says "A user with 15 edits", but plenty more than 15 are listed - in fact there are 803 edits credited to Redrose64 on German Wikipedia, due to their love of importing our templates, modules and indeed articles. --Redrose64 🌹 (talk) 11:29, 29 August 2023 (UTC)
This is a {{Pie chart}}, so that means the developers are your cohorts :). I am somewhat skeptical it can be fixed in a way which works for you and everyone else given what it's doing (Firefox for mobile is not a high-pageview browser). Izno (talk) 23:14, 16 August 2023 (UTC)
Call me a traditionalist, but nowadays I use a Chromebook, and I also have an antideluvian android mobile (now on Android version 10) that I bought second hand several years ago. Both show the pie chart with properly pie shaped slices. So it may just be the fennec browser? ϢereSpielChequers07:08, 17 August 2023 (UTC)
Looking at the rendered html of that graph makes me want to cry. It's impressive as a hack, sure, but even if it is valid css, it's fragile and opaque; surely the Proper Way To Do This is an svg. —Cryptic08:53, 17 August 2023 (UTC)
That would be good but then the chart would quickly become obsolete as it's much harder to update a diagram than some fairly clear wikitext. I know svg can be "easily" updated but a comparatively tiny number of editors would feel comfortable doing that and then going through the image upload procedure. Johnuniq (talk) 09:30, 17 August 2023 (UTC)
Today the colors would probably be done with conic-gradient. That's fairly new but should degrade more gracefully than this absolute mess of absolutes. About 0.5% of English Wikipedia page views would not support it (0.13% Chrome/Edge, 0.14% IE, 0.21% Safari). That's probably a reasonable percentage to try and improve things. Izno (talk) 17:34, 17 August 2023 (UTC)
Yet another example of a broken pie chart
Comment: I am unarchiving this stale discussion, due to finding another person who experienced a similar bug, @2NumForIce
I tested the page linked above on Firefox for Android 68 (last version supporting some old Android versions; at least I think I did, that was several days ago) and just now version 117, and neither displayed this issue. (I did notice some seams running through the middle horizontally and vertically, but presumably that's a consequence of aforementioned looking at the rendered html of that graph makes me want to cry.)
QuickQuokka, if you type about:support in your browser that has this issue, can you post what it lists under the "graphics" section for (1) "Compositing" and (2) "GPU #1" -> "Description"? LittlePuppers (talk) 04:05, 30 August 2023 (UTC)
In trying to link the writer Dan Berger at Ruchell Magee, I found that a google search on Dan Berger shows google has an image of Dan Berger the writer, linking to the Wikipedia article of Brian D. Berger. Is that a Wikidata mixup and does someone know how to fix it ? (I just moved Dan Berger to Brian D. Berger to make way for Dan Berger.) SandyGeorgia (Talk) 12:05, 30 August 2023 (UTC)
Also, I may have moved the lobbyist to the wrong title; one of the sources refers to him as Brian Daniel Berger (Dan), while the other four refer to him as B. Dan Berger. The problem seems to date to 2018, but I may not have fixed it correctly. Just trying to avoid the problem of google mixing up Dan Berger the writer on black activism with B. Dan Berger the lobbyist. SandyGeorgia (Talk) 12:30, 30 August 2023 (UTC)
Google's knowledge panel has a confusing layout. The image is added by Google and selected by their own algorithm. It doesn't link to the Wikipedia article. They got it from https://jsis.washington.edu/humanrights/people/dan-berger/. The Wikipedia link is at the end of a text paragraph from Wikipedia. All other parts may be completely unrelated to Wikipedia. See Template:HD/GKG for a stock answer to complaints about Google errors. You can give feedback to Google that the image is wrong. You can try to upload an image to Brian D. Berger and maybe Google will use it. Most complaints about wrong images in Google panels have been for cases where the Wikipedia article had no image. PrimeHunter (talk) 13:10, 30 August 2023 (UTC)
FWIW I've checked on the Wikidata side - I think you're right that it was involved, as the item for B. Dan Berger was tagged with the Freebase internal ID https://www.google.com/search?kgmid=/m/02vlxwx but Google seems to associate that ID with the historian - everything except the WP article (which it will have picked up from Wikidata) is for Dan Berger the historian.
I've moved that ID across to a new item for the historian, linked to your new page, and I think Google will then pick that up in the near future. Fingers crossed, anyway... Andrew Gray (talk) 18:35, 30 August 2023 (UTC)
That ISBN-fixing code is still useful. On Wikipedia, we write ISBNs as "ISBN ##########", not "ISBN-13 etc." or "ISBN: etc.", so removing that extraneous text is useful. Writing ISBNs in the latter fashions will cause ISBNs to appear on error reports. – Jonesey95 (talk) 06:51, 31 August 2023 (UTC)
Adding images to articles on 'Mobile Edit'
Hello,
This might be a small thing, but is there no way to add images to articles in VISUAL EDIT in mobile editting. As I have checked, the only option is through SOURCE EDITTING. The same is available on PC though, i.e. adding Images or Templates through visual edit.
Okay, follow up question. I see no notification on the talk pages of the articles that use these logos. Is there a way to notify them all? Richard-of-Earth (talk) 14:45, 31 August 2023 (UTC)
In both articles, it looks like the row for Botanic Gardens in the (collapsed) "Station histories" table. I'm not familiar with the template used to populate that row, though. —Cryptic17:26, 31 August 2023 (UTC)
A few more will appear in the category over the next couple of minutes as the template transclusions are processed by the job queue. – Jonesey95 (talk) 17:52, 31 August 2023 (UTC)
Formatting question
What would be the CSS code for blacking out one's User page? I want to do that to mine for a week or two for a private, personal reason. (Well, once I'm told how to do it, I will leave an explanation why. I'm just not interested in broadcasting the reason to the general, disinterested public.) -- llywrch (talk) 22:07, 28 August 2023 (UTC)
I'm not sure what you mean but if you want your own user page to only display the interface for you but display normally for others then you can add this to your CSS:
(Sorry to take so long to respond, been busy with non-Wiki stuff) No, what I'm looking for is a black background -- to the edge of the browser window -- for all to see, & perhaps white type. From your answer, I suspect .mw-body is somehow involved. -- llywrch (talk) 17:38, 31 August 2023 (UTC)
You can only affect the appearance of content within the main content area, by adding style attributes to HTML elements. Thus you could enclose your user page content within a <div style="...">...</div>. To change how someone sees the other parts of the page, you'd have to get them to load a custom CSS file. isaacl (talk) 17:52, 31 August 2023 (UTC)
You could use absolute positioning to cover the whole page, including the interface. But you really, really shouldn't. —Cryptic18:18, 31 August 2023 (UTC)
It sounds disruptive to hide the interface even if it's your own userspace. You can make white text on black background in the main content area with code like: <div style="background:black; color:white; height:100em;">White text</div>. PrimeHunter (talk) 19:44, 31 August 2023 (UTC)
Question - two Page view counts
On Timeless skin sidebar, when I choose Analysis, then Basic statistics. In section "Basic information", "Page views in the past 30 days" - that number is larger than if I page down to bottom and click on "Page view statistics" (External tools section). For example here for Wikiproject Football. Page views in the past 30 days is 1,636; vs Page view statistics is 1,104. Question: which of these is correct/more accurate? Regards, JoeNMLC (talk) 21:48, 30 August 2023 (UTC)
You need to set Page view statistics to 30 days. By default it is 20 days. (I don't know why 20 was picked and it's kind of weird.) Izno (talk) 22:13, 30 August 2023 (UTC)
Thanks - Is there any way to save that 30 day perference? I've tried several things and not able to. (Settings wrench, then tick "Remember chart preference", Save) It always comes up with the 20 days instead. JoeNMLC (talk) 23:22, 30 August 2023 (UTC)
User:PrimeHunter/Pageviews.js adds a "Page views" link under Tools. You could make a modified version with &range=latest-30 inserted in the url. https://pageviews.wmcloud.org actually shows one more day than requested so use &range=latest-29 if you really want 30. I don't know whether it's intentional but latest n in the tool has n days between start and end with both included. The average is computed correctly by dividing with n+1. Your "Analysis" options are added by MoreMenu at Special:Preferences#mw-prefsection-gadgets. The default is to have a "Page information" link to what MoreMenu calls "Basic statistics" in the "Analysis" options. PrimeHunter (talk) 08:22, 31 August 2023 (UTC)
I can change the default range to 30 days. It has been 20 since the demo Mforns (WMF) created many years ago that inspired Pageviews Analysis. There's otherwise no basis to defaulting to 20 days, and I agree going with roughly a month is more sensible. I'll get that deployed soon-ish.
In addition, as PrimeHunter points out, Pageviews Analysis has some logic around date ranges that is admittedly less than ideal. The idea is the past N days should not include today, because it takes up to 24 hours for the pageviews pipeline to populate. The tool incorrectly goes off of your timezone, even though I made efforts to force UTC. So in addition to perhaps confusingly being an inclusive range (the last N days including start and end, where the numerical difference equals N), what you see may not even be the same as another user due to timezones. I'll make another effort to fix that. No one was complaining about that bug here, but this conversation reminded me that it still exists.
I don't know if date ranges inclusive is really a problem or not, but since it was brought up I'll change that as well so that you get exactly N days. Depending on how users react, if at all, this might be changed back to the old behaviour.
Hello, The big orange "You have new messages" bar for logged out users seems to be broken - I can't get it to disappear. I tried editing my talk page by adding a space, but all that did is change the message to "You have new messages from 2 users", even though one of the "messages" was placed by myself. 86.23.109.101 (talk) 19:25, 26 August 2023 (UTC)
@Matma Rex You asked on phabricator if the notifications clear on editing - I've edited my talk page twice and the notification is still there - in fact the notification has updated to say I have messages from 2 users, even though one of them was myself. 86.23.109.101 (talk) 13:07, 29 August 2023 (UTC)
i got a message on my talk page, and saw it thanks to the new notification method. however the notification can not be dismissed, liek the old orange one. this may or not be a big deal but it is something you should know about here at this noticeboard. :^) 173.87.169.10 (talk) 00:16, 27 August 2023 (UTC)
i am not able to reply to matmarex at Phabricator who saidAre you sure that this worked previously? There are bug reports like this going back to 2013 (T59840), and as far as I know, due to the way caching for logged-out users work on Wikimedia wikis, the only reliable way of getting rid of the bar is making an edit. i can only say here that yes, the orange one used to clear when checking new messages, and making an edit doesn't clear it. it is still there. 173.87.169.10 (talk) 23:51, 28 August 2023 (UTC)
Thanks you all for comments, I found the cause of the problem (see Phabricator for details) and I hope we'll get it fixed later this week. Matma Rextalk00:01, 30 August 2023 (UTC)
A couple days ago, I was notified on my phone that someone had posted a message on my talk page with the standard "new message" template. I looked at it and pressed the links given, but the bar wouldn't go away. Two days passed, and it's still there. Today, I checked another phone (one that hasn't been used to edit Wikipedia in over a month), and it has the same template which also will not go away. I then looked at my computer (which technically doesn't even use the same IP address as me), and it has the exact same problem. I am posting here to figure out if this is just me, or a bigger issue, since the notification bar is really annoying and distracting when I am editing or reading the edit history of something. 47.227.95.73 (talk) 02:05, 30 August 2023 (UTC)
I have used my phone, Android v11, for editing for many months without problems. Today, 8/31/23, while logged in on the Android, using Chrome browser v92, when I attempted to edit an article, I got a message that I was blocked by "ST47ProxyBot". The block message said, "The IP address you are currently using has been blocked because it is believed to be used by a peer-to-peer proxy service." The block applies to everything except my own Talk page. The mobile is blocked from editing articles, other Talk pages and my own user page. My connection from my desktop machine, Win10, Firefox v117, is not blocked for editing and I am able to edit normally. The phone and desktop machine both use the same wifi signal, and presumably the same ip address, which is provided by T-Mobile home internet. The block apparently is scheduled to end after only two days. Can anyone explain this state of affairs, and whether I may encounter the problem again? Is this a situation in which my ip address can receive "block exemption"? DonFB (talk) 06:35, 1 September 2023 (UTC)
@DonFB: The short answer is that your mobile and desktop are not using the same IP addresses. Yes this is the type of situation we can use IPBE for your account. I'd consider this a temporary situation and relatively rare. Your phone and the proxy are both likely to change addresses often, making another meeting of the two unlikely. I'll grant IPBE for a month anyway. If you encounter this again feel free to ping me directly. -- zzuuzz(talk)08:12, 1 September 2023 (UTC)
Is it just me or is something wrong with archive.ph? It keeps serving me endless reCaptchas through Cloudflare; I'm stuck beyond it. After a recaptcha it refreshes and has me redo yet another recaptcha. Aaron Liu (talk) 01:20, 2 September 2023 (UTC)
... don't do that? Templates are literally made to be reused. LST is a mechanism that you should employ almost exclusively in the mainspace, and I can think of no reason to use it in template space. Izno (talk) 19:16, 31 August 2023 (UTC)
@Jarble: It shouldn't affect the behaviour whether pages are in template space but it's unclear what you tried to do. If you give a saved example and say what you wanted then we can probably say how to do it or give another way to get a similar result. PrimeHunter (talk) 19:26, 31 August 2023 (UTC)
@PrimeHunter: I found several templates that include labeled sections. I tried to transclude one of these sections here, but nothing is transcluded:
I get an error message when I click on the URL in the ref but if I follow the directions given in the discussion above (directions were first given here), I get to the same place, and copying the URL doesn't change it from what it was before.
No, to get to the URL you have to follow a very specific process, and only then do you see the exact same URL at the top of the screen. Merely clicking on the URL gets you an error message.— Vchimpanzee • talk • contributions • 23:04, 1 September 2023 (UTC)
When I click on the link provided by LuckyLouie, which is like the one you provided except with https, it works for me. Are you having difficulty accessing it directly? isaacl (talk) 00:07, 2 September 2023 (UTC)
Up front, I hope this doesn't come off as crass. When I see a new entry at WP:Deceased Wikipedians, since full protection is SOP and I can edit fully protected pages I try to help straighten out any loose ends. Today I saw the news about Nosebagbear, and as per the guidelines (and yes, I'm the one who added this bit) I added the nocat=yes parameter to a couple userboxes in these edits, but it doesn't seem to have removed the account from the relevant categories. The only other solution I know of is to subst those userboxes and manually remove the category parameter, which seems way too complicated to be the only alternative. Is there an issue with the userboxes, the parameter, or something taking up space between the chair and the screen? The Blade of the Northern Lights (話して下さい) 02:58, 2 September 2023 (UTC)
Thanks. When I saw adding that parameter didn't have the desired effect, I thought it best to stop messing around with anything else until it was clear what was going on. I'm happy to take care of the rest. The Blade of the Northern Lights (話して下さい) 03:08, 2 September 2023 (UTC)
I think I've gotten all the rest (I'm not worried about the inclusionist Wikipedians one). Thank you. And as long as I'm here, I raised a glass to his memory and encourage anyone else here to do the same. The Blade of the Northern Lights (話して下さい) 03:13, 2 September 2023 (UTC)
to keep this short, i am currently writing this from an amazon kindle (reading device) and attempting to access Wikipedia from it is slow, buggy,and prone to crashes, it would be awesome if this could be dealt with as soon as possible — Preceding unsigned comment added by 86.10.35.182 (talk) 21:35, 25 August 2023 (UTC)
There's a growing feeling (not sure I could actually call it a consensus yet) at WP:VPP#Proposal to remove "rearrangement of text" from definition of minor edit. that we should deprecate the whole minor edit feature. Let's assume for the sake of argument that we were going to do that. What would it look like from the technical point of view?
There's a rev_minor_edit field in the database. We certainly wouldn't remove that field; enwiki would just stop using it. There's a "This is a minor edit" checkbox on the edit form. Would would it take to make those go away? I'm assuming that could be done by an interface admin, at least for the web-based edit surfaces. And I assume the mobile apps would need to be modified at the source code level to delete those?
The various APIs all support setting that bit on edits. So I assume we'd need some kind of per-wiki configuration flag to say if rev_minor_edit is supported, and then figure out what to do if you try to set it on a project that doesn't support it. The two obvious choices are "silently ignore it" and "return an API error". I'm not sure which of those would be worse.
Special:Contributions and page histories (and maybe other places?) display an "m" on minor edits. I'm again assuming that's something that can be deleted by just editing the interface? I assume there's various search and filter forms which allow you to only look for minor edits? And of course, gazillions of tools and libraries would need to be modified to be aware of this. What else? I'm sure I've only scratched the surface. RoySmith(talk)14:10, 29 August 2023 (UTC)
In theory, there's a simple config option: $wgGroupPermissions['user']['minoredit'] = false; would deny the right to mark edits as minor to logged-in users, just like it is currently denied to logged-out users. You'd still have to test various editing interfaces to make sure that they support this option – they probably do if they support logged-out editing, but someone might have assumed that logged-in editors always get this option instead of checking the user rights. Matma Rextalk15:03, 29 August 2023 (UTC)
Would probably want to add it back (if it's not in the default set) for bots.
Looks like rollback checks for the minoredit user right. So if that right were removed from users and given to bots, only bot rollbacks would be marked as minor. Anomie⚔23:39, 2 September 2023 (UTC)
We probably shouldn't do that, if nothing else to avoid confusion for all the old revisions in the database that are marked as minor. Removing the ability to flag new edits as minor (as Matma Rex described above) should be sufficient. Anomie⚔23:39, 2 September 2023 (UTC)
My apologies. I did not intend on stepping on any toes. I just noticed an assertion, and was curious about their perspective. Thank you for the link, I'll go check it out : ) - jc3708:39, 30 August 2023 (UTC)
The vandal edit has neither an edit summary nor an editor, but you can still see the vandal's username in the (automatically generated) edit summary of the revert. --Hob Gadling (talk) 05:14, 3 September 2023 (UTC)
hyphen-minus adjacent to opening curly brace problem?
What is is about -{ in a template that causes MediaWiki to return the wikitext instead of the rendered template? Examples:
{{cite book|title=sommat-{and sommat else} and more}} – does not work
{{cite book |title=sommat-{and sommat else} and more}}
{{cite book|title=sommat-{and sommat else} and more}} – works
sommat-{and sommat else} and more.
{{lang|es|sommat-{and sommat else} and more}} – does not work
{{lang|es|sommat-{and sommat else} and more}}
{{lang|es|sommat-{and sommat else} and more}} – works
sommat-{and sommat else} and more
Just playing around I discovered that this works:
{{cite book|title=sommat-{and sommat else}- and more}} – does work
sommat-{and sommat else}- and more.
That suggests that -{...}- means something to MediaWiki. What does MediaWiki think that markup means?
In all of these examples the dash character is U+002D hyphen-minus. I discovered this issue in this reference where the editor got 'round the problem with <nowiki>...</nowiki> tags.
They're markers for automatic language conversion (such as between traditional and simplified Chinese). See mw:Writing systems/Syntax. I expect when the parser sees the opening hyphen-brace without the closing pair, it just gives up (which isn't unreasonable). —Cryptic17:09, 3 September 2023 (UTC)
Thanks. But should that functionality be enabled at en.wiki? I can see it being useful at sr.wiki and sh.wiki where, apparently, the languages have both Cyrillic and Latin forms – the markup is mentioned at Serbian Wikipedia. That does not hold for en.wiki for which there is only one alphabet so it seems inappropriate to enable that markup here. Am I missing something?
-{{ with double brace seems to work. -{ with neither a following { nor matching }- does occur in the wild, especially in drug articles, and seems to be a known problem which editors have worked around in various ways. For example, Pesampator uses -{{(}} and AZD-1940 uses nowiki. Certes (talk) 18:27, 3 September 2023 (UTC)
Are there any cases where -{whatever}- actually needs to be processed in some way other than producing literal text with hyphens and braces (as in some IUPAC names like RAD140: click [show] in infobox)? Certes (talk) 19:48, 3 September 2023 (UTC)
On the iOS (Safari and Firefox) mobile website version of the landing page of wikipedia.org, it does not display the Wikipedia logo but looks like this in portrait mode, showing a cropped version of the WMF logo. However, rotating to landscape mode displays the correct logo. I'm not sure if this is related but it happens on both an iPhone 15 and SE running iOS 16.6 as well as an iPhone running the 17.0 beta. Apologizes if this is already known or being addressed but I've had a couple of people I know reach out to me about it. - Aoidh (talk) 22:33, 3 September 2023 (UTC)
But is it not the case that 'the software' (treading carefully around things I know less than nothing about) already copes with 'incorrectly-formatted' page numbers eg |pp= containing eg hyphens rather than n-dashes? If this is so, is it not possible to silently correct "anything other than n-dash" in |date= in the same manner?
The message CS1 maint: date format (link) is not javascript but is emitted by Module:Citation/CS1 when it renders dates that use hyphen-minus (-) when they should be using an en dash (–). You enabled that message display in your common.css.
MOS:DATES, particularly MOS:DATERANGE, does not allow hyphen separators in any date format except the year initial numeric YYYY-MM-DD format. cs1|2 complies with MOS:DATES. When cs1|2 renders a date that has a hyphen separator, the date is rendered with an en dash. Because the date was 'modified', cs1|2 emits a maintenance message and a category link so that editors can fix the date to bring it into compliance with MOS:DATERANGE.
Because page numbers may legitimately use hyphens (page 21-5 for the fifth page in section 21, for example), cs1|2 accepts hyphens in |page= as hyphenated page numbers and attempts to get it right when hyphens appear in |pages=. Examples:
{{cite book|title=Title |page=1-2}}
Title. p. 1-2. – looks like a hyphenated page number but could be a page range misusing the hyphen separator; cs1|2 can't know
{{cite book|title=Title |pages=1-2}}
Title. pp. 1–2. – looks like a page range but because of the hyphen, could be a hyphenated page number; cs1|2 can't know
{{cite book|title=Title |pages=1-2-1-4}}
Title. pp. 1-2–1-4. – a hyphenated range of hyphenated page numbers?
Because of the complexity of pagination, cs1|2 doesn't emit maintenance messages for |page= and |pages=
@Trappist the monk: I manually fixed the templates that aren't DYK nominations and the Wikipedia pages that aren't Articles for Deletion. I ran a streamlined version of the bot, and the category went from 803 pages to 45 pages). GoingBatty (talk) 21:38, 3 September 2023 (UTC)
Many thanks for the low-down explanation, much appreciated since I have hundreds of pages like 7-2-43 to deal with in sfns. Looks like something worked, anyway. MinorProphet (talk) 22:15, 3 September 2023 (UTC)
I noticed a few days ago that the notification button looks different.
Before it had a red background, making it pop out and be the first thing I notice. Now it's white text on a light grey background (Literally like this).
I don't have any visual impairments, and I can't see it almost at all. I can't imagine how this accesibility monstrosity feels for visually impaired and people with other such disabilities.
it seems that was a glitch. i got the same result on my phone on all 2.5 layouts (but not my pc), but on dark mode. switching to light mode and back fixed it
whatever that was didn't happen again, but to balance that out, the dark and light mode buttons decided they should both appear in the same time and place, and that the one that didn't work should go over the other cogsan • (give me attention) • (see my deeds)17:44, 1 September 2023 (UTC)
@QuickQuokka: Correct, this is just a quick fix until the devs get around to fixing this properly. Also, see WP:VALIDALT. Creating a second account to ping yourself is definitely not sockpuppetry (I have about a dozen such accounts!) and is quicker than waiting for someone to ping you. Suffusion of Yellow (talk) 03:45, 2 September 2023 (UTC)
Is there a way to display a user's gender terms ("Gender used in messages" on Special:Preferences) without enabling Wikipedia:Tools/Navigation popups? I want to use the appropriate pronoun to refer to a user, but I don't like Navigation Popups in general, and turning it on/off just for the purpose of getting someone's gender seems overkill. Mitch Ames (talk) 00:47, 2 September 2023 (UTC)
See Template:They and the others linked from its documentation. {{they|Mitch Ames}} → he. They subst cleanly, or can be used in preview mode if you don't want the templates hanging around. —Cryptic00:55, 2 September 2023 (UTC)
I noticed the other day that we can’t ‘thank’ the bots that make our lives as editors easier. It made me wonder how often the bots would appear in the thanks log if that was not the case. (To be clear, I’m not starting a formal proposal here, just a light-hearted comment )
On desktop, when I put the cursor into the Search box, it immediately opens a dropdown, with the name of a band I looked up, several weeks ago. Sometime over the last few days, this has changed, and now offers me the choice of three bands I've looked up in the last while. How do I turn off this autocomplete feature? I can't find anything in Preferences > Search, or Gadgets. Using Chrome, 116.0.5845.141, 64-bit. TIA, BastunĖġáḍβáś₮ŭŃ!14:25, 4 September 2023 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Recent changes
EditInSequence, a feature that allows users to edit pages faster on Wikisource has been moved to a Beta Feature based on community feedback. To enable it, you can navigate to the beta features tab in Preferences. [44]
There is a new option for Administrators when they are changing the usergroups for a user, to add the user’s user page to their watchlist. This works both via Special:UserRights and via the API. [46]
The LoginNotify extension was not sending notifications since January. It has now been fixed, so going forward, you may see notifications for failed login attempts, and successful login attempts from a new device. [48]
Changes later this week
The new version of MediaWiki will be on test wikis and MediaWiki.org from 5 September. It will be on non-Wikipedia wikis and some Wikipedias from 6 September. It will be on all wikis from 7 September (calendar).
Starting on Wednesday, a new set of Wikipedias will get "Add a link" (Eastern Mari Wikipedia, Maori Wikipedia, Minangkabau Wikipedia, Macedonian Wikipedia, Malayalam Wikipedia, Mongolian Wikipedia, Marathi Wikipedia, Western Mari Wikipedia, Malay Wikipedia, Maltese Wikipedia, Mirandese Wikipedia, Erzya Wikipedia, Mazanderani Wikipedia, Nāhuatl Wikipedia, Neapolitan Wikipedia, Low German Wikipedia, Low Saxon Wikipedia, Nepali Wikipedia, Newari Wikipedia, Norwegian Nynorsk Wikipedia, Novial Wikipedia, N'Ko Wikipedia, Nouormand Wikipedia, Northern Sotho Wikipedia, Navajo Wikipedia, Nyanja Wikipedia, Occitan Wikipedia, Livvi-Karelian Wikipedia, Oromo Wikipedia, Oriya Wikipedia, Ossetic Wikipedia, Punjabi Wikipedia, Pangasinan Wikipedia, Pampanga Wikipedia, Papiamento Wikipedia, Picard Wikipedia, Pennsylvania German Wikipedia, Palatine German Wikipedia, Norfuk / Pitkern Wikipedia, Piedmontese Wikipedia, Western Punjabi Wikipedia, Pontic Wikipedia, Pashto Wikipedia). This is part of the progressive deployment of this tool to more Wikipedias. The communities can configure how this feature works locally. [49][50]
In the past few weeks, I've been getting the following error message whenever trying to visit Special:TopicSubscriptions (also see screenshot):
Fatal exception of type "Wikimedia\Assert\ParameterTypeException"
I've checked my browser's console, and the following error shows up (it doesn't show on any other wiki page besides Special:TopicSubscriptions):
Failed to load resource: the server responded with a status of 500 ()
This bug doesn't occur on other projects (e.g. Wiktionary), nor does it occur when I use my sock puppet. I therefore suspect that it has something to do with the fact that I'm subscribed to a relatively big number of topics, probably more than 200.
You should file a report on Phabricator. If you do, it's important to include the random string of hex digits in the error message (starting with "d22f8..." in your picture, but will be different every time), as that allows to developers to determine the exact cause of the error. * Pppery *it has begun...22:22, 5 September 2023 (UTC)
Thank you. I actually looked for something similar on Phabricator, and was pretty surprised that no one appears to have reported anything like this. I will be filing a report tomorrow, but if anyone recalls seeing this error message on that specific special page, I'd like to hear about that just to know I'm not alone ;-) Guycn2 (talk) 22:26, 5 September 2023 (UTC)
@Vchimpanzee: At {{If IP}}(the template used in that talk page) do you see both messages?
The template uses -show classes, so both the logged out and the logged in messages are there in the page, it's just that either one or the other is hidden(using CSS) if you aren't part of that user group - so technically if you are seeing both messages you could work around this by adding to your common.css the line .anonymous-show { display: none !important; } (to mirror what should have been set automatically at MediaWiki:Group-user.css).
Do you remember if you ever made any personal CSS or Javascript changes to enable seeing messages for anonymous users? I didn't see anything in your common.css or common.js file that seems to affect this, but perhaps you have a different customization file somewhere, or you set up globally-used files on mediawiki.org? isaacl (talk) 16:29, 5 September 2023 (UTC)
You are not logged in.You are currently logged in.
I only see "You are currently logged in", as intended. If you also or only see "You are not logged in" then which device or browser are you using to view this page? If it's a Windows browser then does it make a difference to press Ctrl+F5? PrimeHunter (talk) 20:01, 5 September 2023 (UTC)
You are not logged in.
I only see "You are currently logged in", as intended. If you also or only see "You are not logged in" then which device or browser are you using to view this page? If it's a Windows browser then does it make a difference to press Ctrl+F5? PrimeHunter (talk) 20:01, 5 September 2023 (UTC)
My keyboard has bunch of symbols up there and even with my new bifocals the only one I can see clearly enough to recognize is a magnifying glass. It's the ninth key to the right of ESC.
If it's a laptop, you might need to press the Fn key (like a shift key) at the same time, to get the F5 function (in which case "F5" might be written very small on the key - see top of File:Partial keyboard.jpg for example with F7, F8 ). It depends on your specific machine's configuration. Mitch Ames (talk) 23:59, 5 September 2023 (UTC)
Not a laptop. Anyway, we're trying to solve an unrelated problem, and the original problem is one I don't even care that much about, but we have possible solutions here for that.— Vchimpanzee • talk • contributions • 19:57, 6 September 2023 (UTC)
I see you have a meta:User:Vchimpanzee/global.css file, but I don't believe anything in it would cause messages for non-logged in users to display. The only other thing I can think of is somehow MediaWiki:Group-user.css never gets loaded for you, but I don't know what might be causing that for your case (it should be loaded for all logged-in users). Have you tried logging out and logging back in again? Maybe there's some server state that's out of synch. isaacl (talk) 17:54, 7 September 2023 (UTC)
When one hovers over certain links, a popup shows some initial text and an image. Here are some examples:
religious terrorism, religious violence and terrorism.
The religious terrorism image is File:Lipoma 04.jpg, which documents a surgical procedure and has nothing to do with the topic except perhaps to suggest dismemberment. The image selection seems to be part of the hovering mechanism, but I haven't been able to figure out where that mechanism gets images. I could only find the image on Commons by asking my browser to copy the image link. The ability to include images in hover-over popups without any review has obvious NPOV issues. How does this mechanism work, who decides which images are appropriate and how can choices be challenged?--agr (talk) 16:48, 7 September 2023 (UTC)
Most puzzling, especially puzzling as the images for religious violence and terrorism make sense while the one for religious terrorism does not. Horse Eye's Back (talk) 17:49, 7 September 2023 (UTC)
It seems to be less targeted than that. I came across this while looking at Category:Numeral systems. I thought it was targeting South Asian languages at first but Etruscan numerals and Georgian numerals are also affected. The documentation for popups states that images are sourced through the PageImage Extension, which claims the associated image can be seen in the page information, but all the pages I've checked dont have an image listed there at all. I suspect someone may have hidden it in a common template but I don't know if that's even possible. It's definitely a serious exploit though. krytton (a) 17:59, 7 September 2023 (UTC)
The image is chosen by mw:Extension:PageImages#Image choice. A vandalised template in the lead can cause a bad page image. The preview is made by mw:Page Previews which uses caching, so a former page image may be displayed in previews for a while. "Page information" shows the current page image which may be different from one still used in previews. PrimeHunter (talk) 18:56, 7 September 2023 (UTC)
I feel sure this is a known bug but I couldn't find it on Phabricator. The entry above appears on my watchlist. The problem is that the link to en:Starý most actually points to wikidata:Starý most which does not exist. Similar with the link to en:Starý most (Bratislava) — Martin (MSGJ · talk) 17:16, 6 September 2023 (UTC)
I can reproduce it by watching Starý most (Bratislava) (watch the redirect, not the target) and enabling "Show Wikidata edits in your watchlist" at Special:Preferences#mw-prefsection-watchlist. This is phab:T186629 from 2018, the first search result on "Language link changed from". The original edit was the local move [51]. It triggered a Wikidata edit [52] which created the watchlist entry with wrong links. The HTML for the links:
<span class="comment">(Language link changed from <a class="external" href="https://www.wikidata.org/wiki/Star%C3%BD_most_(Bratislava)">en:Starý most (Bratislava)</a> to <a class="external" href="https://www.wikidata.org/wiki/Star%C3%BD_most">en:Starý most</a>)</span>
The same error happens for other watchlist entries about Wikidata edits caused by moves at the English Wikipedia. If a move was at another Wikipedia language then the links go correctly to that wiki.
For me the watchlist entry at least gives the correct name "Starý most (Bratislava) (Q435423)" for wikidata:Q435423. Your screenshot says "List of crossings of the Danube (Q435423)" which appears to be a separate error. "List of crossings of the Danube" is wikidata:Q911165 which hasn't been edited since June. PrimeHunter (talk) 19:52, 6 September 2023 (UTC)
Yes, List of crossings of the Danube is correct because I am watching that article and it uses many entities including Starý most (Q435423). T186629 looks like the right bug, well done for finding it. So the only question is why it takes 5 years to fix something apparently so simple. — Martin (MSGJ · talk) 21:24, 6 September 2023 (UTC)
A bug with OpenStreetMap maps in Wikipedia Infoboxes
Hi, in the case that coordinate of a place is entered inside the Infobox of an article, then incorrectly red colored area indicator of that place is not shown in the map of that Infobox. For example, if we place coordinates of National Zoo of Malaysia inside its Infobox, the rendered map would be:
Code
Rendered Infobox
While if we place that coordinates outside of the Infobox (at the top), then the red colored area indicator would appear in the map, correctly. Like this:
This does not appear to be a bug. Adding |mapframe-wikidata=, as described in the template's documentation, appears to work fine. – Jonesey95 (talk) 17:34, 8 September 2023 (UTC)
@Jonesey95 With the parameter |mapframe-wikidata = yes the Infobox area indicators appears, but we are talking about default scenario. If the bottom code by default does show the area indicators, then the top code by default should show them too, or the inverse default scenario should be applied.
I think in 99% of cases these red line area indicators are necessary, so the default value of mapframe-wikidata should be |mapframe-wikidata = yes, but the user could change this parameter to no for 1% of cases. Hooman Mallahzadeh (talk) 03:31, 9 September 2023 (UTC)
Looks like this is a bug with the way the links show in MediaViewer; they seem to work just fine if you click on it in the standard view. — xaosfluxTalk18:56, 9 September 2023 (UTC)
Happy Thursday! Some of these pencil icons are pretty big now! On the Refdesks, the "edit" icon for the date sections seem like they're scaled 2x the individual questions. Meanwhile I had a look at a mainspace article and the icons to edit lvl3s are larger than the ones to edit lvl2? Then lvl4s and lvl5s are back down to normal size (see History of China for an example lvl5 subheadings in the wild). I don't think I have anything turned on that would affect this. Minerva, Firefox, Android, if any of that matters. Just some brief feedback. Folly Mox (talk) 07:50, 7 September 2023 (UTC)
Hmm I do note that I'm now unable to edit AfDs by tapping the pencil icon from the daily transclusion log. It just recollapses the AfD entry, so now I have to go to the article, open the AfD notice, then click through to the individual discussion before I can edit (unless I'm using ReplyTool, which is not always possible, for example after a relist). Pretty sure this behaviour isn't intended, and I'm pretty sure it started the same time the pencils became engorged. Folly Mox (talk) 18:53, 10 September 2023 (UTC)
The WMF may be planning a performance budget for default gadgets
Some say I shouldn't worry and there's no need to inform the community at this point. But my belief is that the community can provide valuable suggestions, alternatives and ideas.
“
Currently anyone with the correct rights can edit MediaWiki:Gadgets-definition to turn gadgets on for anonymous users. Left unchecked this has led to some performance problems and potential SEO (related to slow loading) on many of our smaller sites leading to phab:T340705. We should be doing more to guide site admins on when they are introducing performance/SEO problems to the site. In MediaWiki we've been investing a lot of effort in limiting the amount of CSS and JS we ship to our end users, but any improvements here are meaningless if the same is not happening within gadgets. I think we should enforce a performance budget for all gadgets that have been marked as "default".
The ticket does point out that enwiki has less than a 1% increase when community gadget/site scripts are shipped. I hope that nothing we run here is covered, but perhaps we can ensure that message gets through and we at least get notification of any breakage which may be imposed on us. Fortunately, our HTML has very little of the advertising which would hit response times. Although I don't share the poster's concern with SEO, if we must regard other sources of information as rivals rather than partners then the lack of ads should also help us to rank ahead of them. Certes (talk) 22:53, 10 September 2023 (UTC)
There appears to be an issue with Template:Advert not displaying in Firefox (my version is 117.0). The page does not load and only displays a white blank page. This also means any pages with the template do not display the template while the rest of the page does display. Osarius09:01, 8 September 2023 (UTC)
Not experiencing this on Firefox 117.0 on macOS. Could it be an extension, maybe an adblock cosmetic filter that filtered it as a likely advert? DFlhb (talk) 09:16, 8 September 2023 (UTC)
Not experiencing this either. I have bunch of extensions enabled among which includes adblockers with bunch of filters enabled (none are cosmetic) and trackers blockers, and Firefox built-in Enhanced Tracking Protection enabled in Custom mode. If you haven't done so, try launching Firefox in Safe mode and visit the Template:Advert and see if it's displaying. —Paper9oll(🔔 • 📝)09:59, 8 September 2023 (UTC)
I see the template as normal on Firefox 117.0. However, I've whitelisted Wikipedia in my ad blockers. I expect one of your extensions thinks that a page called Advert might be an advert. Certes (talk) 11:14, 8 September 2023 (UTC)
It's partially published on toolforge (see User:GreenC bot/Job 15), but parts aren't world-readable so I can't test. (Also, the last time I used awk for more than a few minutes at a time was before unicode existed.)I suspect the problem is the -b flag in fambot.awk; it looks to be what's preventing the "β-Hydroxy β-methylbutyric acid" in WP:Featured articles from being properly upcased to "Β-Hydroxy β-methylbutyric acid" as fetched from the category:
Hmmm that's weird as I had it working with a static override in the bot code for this particular name. Something in the past few days changed somewhere, it wasn't my code. The -b is interesting, I could try removing it, it might break other things. This has used up a lot of time. I keep trying. -- GreenC16:09, 11 September 2023 (UTC)
User:SandyGeorgia: This is the problem. Recall we added this workaround so the bot sees it as B and not β since the Category page returns it as B we want WP:Featured article to also return it as B. I'm hesitant to try Cryptic's suggestion of removing the -b switch because it might break other things, but, I can try that if you prefer, the worse that will happen is the report has false positives and we add the -b back. -- GreenC18:34, 11 September 2023 (UTC)
Updates on Vector 2022, Zebra, dark mode, and more
Hi everyone,
First of all, many thanks for your involvement in discussions about the desktop interface and the Vector 2022 skin. Below, we give you an update on Vector 2022 and Zebra, as well as our next project, which includes introducing dark mode, and customizations on mobile and desktop for both logged-in and logged-out. Let's do this!
Last few months: Zebra, preference persistence, annual plan, and more
During our conversations about Vector 2022, we mentioned that we see the skin as an improvement in itself, but also, as a foundation on top of which we can continue to make improvements. We also discussed ideas for changing our default interfaces on desktop and mobile browsers.
Over the last couple of months, we have continued working on skins through bug-fixing, performing an A/B test on Zebra (read the new update with results from the test!), and studying our previous user testing on font size. We have also taken a close look at the latest Community Wishlist Survey's top #1 proposal, dark mode, as well as other accessibility-related requests from communities. Those of you who have read the annual plan or took part in the consultations of its draft, might have also noticed this related objective (WE2.1). If you watch our team even more closely, you may have read our May update where we captured some of the initiatives mentioned above.
In addition to that, we have worked on building out the capability for logged-out users to customize certain portions of the interface. First of these were introduced as a result of the conversations we had together around customizing the width of content in Vector 2022. We are glad your community encouraged us to build a way for readers to set these options and look forward to using it for changes that improve accessibility!
Next project: Improved typography and dark mode
Now, we are ready to work on a next project. Our focus will be on making the site more accessible while reading and easier to customize. We will work on:
Providing more options for font size and typography (currently, these are only available on mobile) and improving our default typography
Building out a dark mode
Both of these will be available on desktop and mobile, and customizable for all logged-out and logged-in users.
We have published our initial thoughts on this project page, as well as some of our initial research and review of existing literature for the project. We welcome you to read over our thinking and share your opinions below. We are currently defining what success means for this project. Sketching out summaries of the research we’ve done so far that made us realize that accessibility should be a priority for this work.
We are also building a tool that will allow editors to experiment with different options around font size and typography. The goal here would be to identify what needs different communities have for various customizations and make it easier to discuss ideas for the changes together. We hope to have this tool ready to share with you all within the next couple of weeks.
Thanks for the update. It looks like Wikipedia:Vector 2022 could use some updates. It doesn't show that anything has happened since January 2023.
Also, is the team still processing bug reports? Things like T325219, an easy fix that would eliminate some "oops" whitespace at the top of every single page, sitting idle since February do not inspire a lot of confidence. Other feature requests sit idle, like T333590, asking for notification icons to be in the sticky header, and T329673, asking for the sticky header to appear in all pages and page views (e.g. diffs and history). I've had to work around these basic problems since December using personal CSS. Editors should not be expected to have to do this with the default skin. – Jonesey95 (talk) 03:12, 12 September 2023 (UTC)
I don't process Wikimedia bug reports, yet for T333590, all I see is a personal assertion of editor behavior that is not backed up with data. As a neutral observer, this bug is no more important than any other random feature request. If there is evidence that most editors expect certain behavior, I suggest including the data for that so it can be properly evaluated. Without that, it sounds like a personal wish list item. Orange Suede Sofa (talk) 07:38, 12 September 2023 (UTC)
I think it was unintentional… <noinclude> tags were first added with a bot notice: [53]. Then another user, probably accidentally, moved a hatnote inside these tags: [54]. The bot notice was later removed: [55] but the tags were not. (They could also be used legitimately for article excerpts, but that doesn't seem to be the case here.) Matma Rextalk15:05, 11 September 2023 (UTC)
That article has been contentious, so I'm a bit reluctant to fix this myself without ample feedback. Does anyone else see any reason for the noincludes? SandyGeorgia (Talk) 16:52, 11 September 2023 (UTC)
Many talk pages appear to transclude their corresponding article. Some templates in the talk banners, such as {{Talk header}}, will be referencing the page in a way that causes a transclusion to be recorded. For example it might run getContent() on it in a Lua module to determine what sort of page it is (article, redirect, dab...). Certes (talk) 18:09, 11 September 2023 (UTC)
When an article transcludes itself, 99+% of the time it uses {{cite xxx}} templates (Trappist the monk has explained why in the past). The other tiny fraction are genuine errors, this probably isn't such a case. As for talk pages transcluding their articles - {{talk header}} and {{WikiProject banner shell}} both do this, as do all WikiProject banners. --Redrose64 🌹 (talk) 13:30, 12 September 2023 (UTC)
not only aren't they indented right a couple of characters, or even flush with text above it, they are
actually shifted to the left, as if there were a bit of negative left padding. This seems very odd.
What could be going on here? Is there a ticket for this already, or should I write one?
Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum. Mathglot (talk) 08:29, 10 September 2023 (UTC)
That's just how the CSS works out. The margin is on the <ul>, but that extends under the image as the floated image only pushes the text right. There's a template {{flowlist}} that can wrap the list to avoid this display issue. That's also why MOS:IMAGELOC suggests avoiding left-floated images when bulleted lists are nearby. Anomie⚔11:53, 10 September 2023 (UTC)
This is a known issue, and has been so for as long as I've been around (14+ years). That said, it's improved a little during that time - when I first started, the bullets actually overlapped the image. --Redrose64 🌹 (talk) 18:54, 10 September 2023 (UTC)
Thanks, all. Just to be clear, when we say that this is "just how the CSS works out", I assume we mean that "this is just how the Mediawiki CSS works out". Given the CSS box model, I'm assuming that if I tried this in a pure Html file external to Wikipedia using an html unordered list and rendered it in a browser, the indentation would be correct. In other words, this is a Mediawiki bug, not a bug in CSS. I'm not pushing for it to be resolved if it's been outstanding this long, I just want to be clear where the locus of the problem is, in case this comes up again. Mathglot (talk) 19:49, 10 September 2023 (UTC)
you will find that the first few list items (six in my browser) have their bullets overlapping the image, with the left of the letter "L" immediately to the right of the right-hand edge of the image; and the last few list items (four in my browser) are below the image, arranged normally. So you need extra CSS to move those first few bullets to the right. --Redrose64 🌹 (talk) 21:30, 10 September 2023 (UTC)
Also six in my browser. Well, I'm stunned. So the problem is with CSS itself, then, as I can't see any way that this is compliant with the box model, unless there's some very fine print somewhere. Wow. So the solution here is to use {{flowlist}}; thanks for that great tip, and all the supporting work to demonstrate this. I'm still getting over my shock. I wonder what else is lurking out there in CSS-land that I don't know about. (Don't tell me; I don't think I really want to know!) Mathglot (talk) 22:39, 10 September 2023 (UTC)
Floated elements are removed from the layout flow and so the bounding boxes of other elements don't take them into account. However floated elements do push aside the inline content of the other elements. Floats have always been tricky to use as a result—you often have to allocate space for them within other elements—and thus other layout methods like grid or flexbox were invented and are usually preferred nowadays. isaacl (talk) 00:31, 11 September 2023 (UTC)
@Isaacl:, thanks for this. My skills are mostly a robust 4.1, with some of the bright and shiny 5.0 stuff, but I'm not up on things like grid or flexbox, so at least I know where to look to update my increasingly rusty skills. Much appreciated. Mathglot (talk) 01:04, 11 September 2023 (UTC)
I think the "fine print" you're missing is that the list item bullets ("markers" in CSS terms) are not a part of the box model. They are drawn outside of the boxes of the list items, regardless of what else appears there. And by the way, there's a workaround in CSS that changes that – list-style-position: inside[56] causes the markers to be placed inside the list item's box, and behave more intuitively. The drawback is that you can't adjust where exactly the marker appears using margin and padding any more. Matma Rextalk14:50, 11 September 2023 (UTC)
On a fine technical point, note the bullet will get pushed over by the floated element in that case because it is inline content (though of course that necessarily means it's put within the box content). isaacl (talk) 17:01, 11 September 2023 (UTC)
On another fine technical point, the only actual CSS in my example is the declaration float:left inside the single style= attribute of the img element. The list is pure HTML, using no CSS at all. I know of no non-CSS method for flosting an object such as an image. --Redrose64 🌹 (talk) 19:35, 11 September 2023 (UTC)
<img src=... align="left"> is deprecated Html, but still works using your example, modified; but bullet placement is the same. Mathglot (talk) 21:06, 11 September 2023 (UTC)
The align= attribute isn't deprecated, it's obsolete, which is why I didn't mention it. It's still supported by all major browsers though, but this is not guaranteed. --Redrose64 🌹 (talk) 17:01, 12 September 2023 (UTC)
Noticeboard archive problem - Multiple threads don't display, but material is in the archive
I am having a problem with material in a noticeboard archive that cannot be displayed by viewing the archive, and can only be seen by editing the archive. The material is, in effect, hidden, and I don't know why. Either I am making some complicated mistake, or the viewing of an archive is making some complicated mistake. Is someone knowledgeable willing to take a look at a noticeboard archive? Is there some character sequence that says to hide everything until an unhide sequence is encountered?
I briefly thought that under some circumstances, User:Lowercase sigmabot III was dropping portions of archived threads when adding new threads to a noticeboard archive. After some investigation, I have concluded that, in at least one case, portions of threads are invisible in the archive, although the material that cannot be seen on viewing can nonetheless be seen when editing the archive. The problematical archive file is Wikipedia:Dispute_resolution_noticeboard/Archive_230. I become aware of a problem because I was unable to find a dispute over Purdue University Global. If I view the archive file, I don't see a Purdue dispute. On closer examination I see that thread begins as a dispute about Peter Eckstein-Kovacs and then picks up as a dispute over Theanine. The end of the EK case and the beginning of the Theanine case are not seen. However, if I edit the archive file, I see the Arvind Kejriwal case, then the Purdue University Global case, then Yamam, then Vehicle registration plates of New York, then Vurg, then Theanine.
The problem happens the same if I am using Firefox or Google Chrome or Opera as the web browser. Am I explaining clearly enough what the issue is?
Robert McClenon (talk) 04:38, 13 September 2023 (UTC)
I've fixed it (and then Nardog fixed my fix). You were on the right track comparing the wikitext to the rendered version; you just stopped narrowing it down too soon: it wasn't ==Arvind Kejriwal==, but ===Third statement by moderator (Eckstein-Kovacs)===, that was the first header not displayed, so the problem was just before it. —Cryptic04:52, 13 September 2023 (UTC)
Thank you, User:Cryptic. So there was a loose tag that ate five cases. That sounds like a blurb for a bad horror movie. So I wasn't making a complicated mistake, and there wasn't a complicated mistake in viewing, but there was a complicated mistake in what was being archived. It shuffled along like a zombie until it stopped for some tea, and then the caffeine and theanine got it to wake up. I saw the reference error, but I didn't realize that the reference error would do something like that. Strange. Thank you. Robert McClenon (talk) 05:51, 13 September 2023 (UTC)
Follow-Up Question
So am I correct that this can happen to any noticeboard archive under the wrong circumstances if there is a malformed reference or certain other errors?
Robert McClenon (talk) 06:56, 13 September 2023 (UTC)
Not just that; it can happen on any page, not just noticeboards or archives. An unclosed template or tag has the potential to swallow a large blob of content or even the rest of the page. – Jonesey95 (talk) 13:29, 13 September 2023 (UTC)
Thanks, 163, but now I get the error message: "No archive counter was detected on this page, so archiving was aborted. See User:Equazcion/OneClickArchiver for details." SandyGeorgia (Talk) 16:28, 13 September 2023 (UTC)
Line 47 of the script searches for '\\|counter ?= ?(\\d+)', with no space before the word "counter" and at most one space before and after the equals sign. So at Talk:Operation Gideon (2020) the setting in the "User:MiszaBot/config" is not understood, whereas the one in the deleted "Archive basics" was accepted.
I think the original problem was a mismatch between lines 56 and 74 of the script. Line 56 accepts an optional space between the pipe and the word "archive", but when line 74 does some further processing, it doesn't allow for that space. Thus the text | archive = remains in the name of the archive that the script tries to create.
A workaround for now might be to restore the "Archive basics", placing it above the "User:MiszaBot/config" so that line 56 of the script finds the "archive=" line from the "Archive basics" and not the one from the "User:MiszaBot/config". -- John of Reading (talk) 16:58, 13 September 2023 (UTC)
John of Reading thanks, it worked! Unfortunately, I corrupted the test by moving from my iPad to real computer (different browser) in the interim, so I will next go back and undo and re-test also from my iPad, to make sure it wasn't a browser issue. Thanks again ! SandyGeorgia (Talk) 17:09, 13 September 2023 (UTC)
Hi there. I'm doing a project for my Deep Learning class this semester where I'm attempting to evaluate the quality of articles using a recurrent neural network. In order to do this, I'm hoping to create a dataset where the article's rating is used as the target feature. I know that I can download a dump of the English Wikipedia very trivially, but I have never used the API before, and I'm unsure if I could use it to connect article ratings (and possibly other metadata) to downloaded articles. I'm aware each article title acts as a unique identifier within its namespace, so would I have to download all of the article talk pages, parse them for the class, and manually associate them with every article? Could there be an easier way using the API? All the best, TheTechnician27(Talk page)20:17, 13 September 2023 (UTC)
@Qwerfjkl: I was actually aware of ORES as a new page patroller, and I was hoping to compare my results against its (obviously its will be better, so it's a good watermark). I'll certainly have a look at and try to figure out the PageAssessments extension. Incidentally, I'm going at this with the assumption I can't just query 6.7 million article ratings for free. What sorts of hurdles do you think I might face here? Either way, thank you so much for sharing this! I would've never found it on my own. TheTechnician27(Talk page)00:45, 14 September 2023 (UTC)
@SD0001: That makes a good deal of sense. I've since created a Toolforge account and installed MariaDB. So what I understand so far is that I should make paginated queries to a Wiki Replica. I don't understand a ton about databases, but I think I should be able to figure it out. Regarding the paginations, do you know roughly how often I should query? I know I won't overwhelm the WMF's infrastructure or anything, but I also don't want to be greedy. Additionally, does a school project (possibly attached to an academic paper) qualify as "benefit[ting] the Wikimedia movement"? If I learn more about Toolforge, I might be able to create useful tools in the future, but I'm not under the impression my project RNN could possibly be better than what the WMF already uses. TheTechnician27(Talk page)19:16, 14 September 2023 (UTC)
You can send a new query as soon as the previous query is over. Any research on Wikipedia content quality (or ways to assess it) qualifies as benefitting the movement. – SD0001 (talk) 03:31, 15 September 2023 (UTC)
Mobile web talk pages
The mobile website (idk about the mobile app, I don't use it) talk pages get unreadable if there's too many replies, because it squeezes it literally into one character width column. It is a little bit of a hard problem, because you need to make it clear who's replying to who. Maybe just make it stop getting thinner once it gets to a certain width
Yes, this is a known problem of the current discussion/talk page model. Not really an easy way around it. It’s also relatively uncommon however. —TheDJ (talk • contribs) 09:49, 15 September 2023 (UTC)
"Inline" thing
What on earth is this "Inline" thing that now (as of today) appears on diffs, just below the revision slider and against the right margin? It's got a smartphone-style slidy button. If I hover my mouse over it, there's no hint at all as to what it does. I shall not click the button if I don't know what it does. --Redrose64 🌹 (talk) 21:19, 14 September 2023 (UTC)
This changed yesterday and I thought it may be fixed overnight. Watchlisted items, when clicked on "Differences" to see what has been changed, comes up as a jumble of words and colors mixed with coding. This occurs on the Monobook skin, I haven't checked Vector to read what it looks like. Thanks if coders can bring the "changes" back to the good ole days (Wednesday). Randy Kryn (talk) 13:00, 15 September 2023 (UTC)
In the diff view, do you see a button at the top right that says "Inline"? Have you tried clicking that to see if things go back to normal? –Novem Linguae (talk) 13:20, 15 September 2023 (UTC)
Thanks Novem Linguae and 130.111.39.47, that worked. Yes, I am extremely dumb when it comes to coding and the variations of computer tech, which is why I came here and was quickly set on the right path. My go-to blame is Windows10, which clicks and brings pop ups so often when the cursor is moved around that it seems natural now. Windows7, we hardly knew ye. Thanks again. Randy Kryn (talk) 13:30, 15 September 2023 (UTC)
Hey Randy Kryn, I'm glad to see your query has been answered by Novem above — I do have a quick question though, if you have a moment; is it possible you accidentally toggled that "inline" slider to the "on" (the right hand) position? Only reason I ask is because, as far as I'm aware, it should be "off" by default and we'd ideally like to know if some folx are seeing it enabled by default..! — TheresNoTime-WMF (talk • they/them) 14:08, 15 September 2023 (UTC)
Yes TheresNoTime-WMF, exactly. Either with my thumb in gliding down or the cursor bouncing around. I didn't even know there was such a switch to toggle. So it's not a default problem but a myfault personal glitch. Randy Kryn (talk) 14:22, 15 September 2023 (UTC)
I have tried entering the URL https://en.wikipedia.org/wiki/Gdynia , without logging in, on Chrome, Firefox, Edge, and Old IE. However, on Firefox, Edge, and Chrome, it showed an all-black page without selectable text below header, like https://pasteboard.co/YEpZrWcORKf9.png , while on Old IE, it show a giant Israel flag with no text. The issue would disappear after logging in. THe issue do not happen on other pages like Gdansk. Why is that? C933103 (talk) 10:12, 15 September 2023 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Recent changes
ORES, the revision evaluation service, is now using a new open-source infrastructure on all wikis except for English Wikipedia and Wikidata. These two will follow this week. If you notice any unusual results from the Recent Changes filters that are related to ORES (for example, "Contribution quality predictions" and "User intent predictions"), please report them. [57]
When you are logged in on one Wikimedia wiki and visit a different Wikimedia wiki, the system tries to log you in there automatically. This has been unreliable for a long time. You can now visit the login page to make the system try extra hard. If you feel that made logging in better or worse than it used to be, your feedback is appreciated. [58]
Changes later this week
The new version of MediaWiki will be on test wikis and MediaWiki.org from 12 September. It will be on non-Wikipedia wikis and some Wikipedias from 13 September. It will be on all wikis from 14 September (calendar).
As part of the changes for the Better diff handling of paragraph splits wishlist proposal, the inline switch widget in diff pages is being rolled out this week to all wikis. The inline switch will allow viewers to toggle between a unified inline or two-column diff wikitext format. [59]
Future changes
All wikis will be read-only for a few minutes on 20 September. This is planned at 14:00 UTC. More information will be published in Tech News and will also be posted on individual wikis in the coming weeks. [60]
The Enterprise API is launching a new feature called "breaking news". Currently in BETA, this attempts to identify likely "newsworthy" topics as they are currently being written about in any Wikipedia. Your help is requested to improve the accuracy of its detection model, especially on smaller language editions, by recommending templates or identifiable editing patterns. See more information at the documentation page on MediaWiki or the FAQ on Meta.
Additionally, .ve-init-mw-diffPage-diffMode is the CSS class for the visual/wikitext switch if you also don't want to see that. (And .mw-diff-table-prefix is the class for the div that houses them both.) - Purplewowies (talk) 03:36, 15 September 2023 (UTC)
Hi Mitch Ames, I'll pass this feedback on to the rest of the team — I think we did consider a preferences option to show/hide (and/or set a global default, and thus hide the need for it), but feedback in the past suggested that folx dislike "more things" appearing in their preferences so that likely swayed us against the option. Either way, I'll pass it on — TheresNoTime-WMF (talk • they/them) 09:03, 15 September 2023 (UTC)
folx dislike "more things" appearing in their preferences — More or less than they dislike more things appearing on every diff page? Surely I'm not the only one who views diffs far more often than I view my preferences page. Mitch Ames (talk) 11:43, 15 September 2023 (UTC)
I have tried the feature, and I don't like it. It might be useful if it showed the "as displayed" diff (ie with formatted text instead of wiki markup), but as it stands I find it harder to read than the existing "standard" diff or (for some scenarios) wikEdDiff. I realise that many people might like it, but I think wasting screen space for a switch that is visible all the time - even when I'll never use it - is a bad idea. Mitch Ames (talk) 11:56, 15 September 2023 (UTC)
Diff interface with various toggles
@Mitch Ames There's also an option to show the "as displayed" diff, using the "Visual" button that you should find next to the "Inline" toggle (see image on the right for how it should look).
By the way, the inline diff is basically the same thing as wikEdDiff, so if it is harder to read, perhaps there's some improvement to be made. Can you explain why it feels worse to you? Matma Rextalk22:47, 15 September 2023 (UTC)
For my part, I'd be more likely to use the feature if it didn't come with a gigantic, infantilizing button. Desktop is not mobile. With the visual/wikitext toggle, at least we can hide the redundant icon and debuttonize, but with this, if the immense graphic is hidden, the remaining label is nonfunctional. —Cryptic14:28, 15 September 2023 (UTC)
@Cryptic I can't say I understand why you feel this way about the button, but you make a good point that the text of the label should also be clickable, and we follow that principle everywhere else throughout MediaWiki. It turns out that it was also designed to work this way here, and it's a tiny mistake that broke it – I proposed a patch at T346132 that will correct it, so this will probably be fixed here on Wikipedia next week. Matma Rextalk23:20, 15 September 2023 (UTC)
Ugh, I "fixed" {{Airntd}} before seeing PrimeHunter's response. Ymblanter, should these other categories be moved to match the outcome of the discussion that led to moving the "attack aircraft" category, or should I undo my template changes? Note: I will be away from Wikipedia for about 12 hours soon. Any template editor is welcome to revert my changes to that template if that is the right thing to do. – Jonesey95 (talk) 14:04, 12 September 2023 (UTC)
Or at least I do not think we currently have consensus for moving all these categories. I am just trying to implement a particular CfD conclusion though, this is not really an area I understand well. Ymblanter (talk) 14:12, 12 September 2023 (UTC)
My change has created a bunch of red-linked categories. I have reverted the change pending a more sophisticated fix, a more comprehensive CFD, or an undo of the initial CFD that seems to be tied to other categories. – Jonesey95 (talk) 23:55, 14 September 2023 (UTC)
I was testing out a change to one of my user scripts and noticed that Wikipedia does not have recent changes patrolling enabled on the wiki. I wonder what would it take to turn this feature on? I picture it would be great for stuff like ORES review where certain edits that are below the threshold of ClueBot NG but in the range where they might need fixing could show up in Recent Changes with the "!" indicating unpatrolled. What are your thoughts? Aasim - Herrscher of Wikis ❄️ 22:05, 13 September 2023 (UTC)
@RudolfRed This is a technical question. I am not saying it is impossible to patrol recent changes. What I am saying is that it is impossible to mark recent changes as "patrolled". I wonder if there might be a reason for this. Aasim - Herrscher of Wikis ❄️ 01:41, 14 September 2023 (UTC)
This link might help: WP:Patrolled revisions. Someone must know more about it than me, but I think you might have to delve into the history of Pending Changes. I seem to vaguely recall some incompatibility, compounded by a lack of investment into development, but I might be imagining that. As you've already found out, and will see from the backlinks, being really clear and precise with terminology is key if you want to move it forward. -- zzuuzz(talk)07:07, 14 September 2023 (UTC)
Yeah, this. I also see as a big project it would be impractical to "mark every revision as patrolled" on behalf of the patrollers. Pending Changes does give something similar to RC patrol in which the best revision can be rated. I do think, though, that there would be some benefit in being able to mark individual or sets of revisions as patrolled.
I know New Page Review uses the same patrolling system, but we can give "rollbackers" the ability to mark individual revisions as patrolled but not new pages. Because those are the two tools a rollbacker needs to embark on RC patrol. Aasim - Herrscher of Wikis ❄️ 16:04, 14 September 2023 (UTC)
Patrols seem to be somewhat equivalent to new page reviews on this wiki. Perhaps that is why this ability is still turned on (I see [Mark this page as reviewed] for new pages if the Page Curation toolbar is closed), but is restricted to editors that have the new page patroller permission. –Novem Linguae (talk) 14:56, 14 September 2023 (UTC)
Short answer: we don't have this enabled - and likely won't because of the highly customized PageTriage extension here - combined with our extremely high volume of revision creations. Also, you don't have "patrol" rights (which do work on "new pages" via pagetriage). — xaosfluxTalk15:04, 14 September 2023 (UTC)
I think this question is about individual edits, rather than creating new pages.
If it seems impractical to "mark every revision as patrolled", then presumably it is even more impractical to have multiple editors look at the same edit several times. Whatamidoing (WMF) (talk) 16:02, 15 September 2023 (UTC)
I think we can also consider that after a certain time the edits are automatically marked as patrolled. I think it is variable and can be changed but if, say, the edits are more than a day or even half a day old, it can probably be removed from the patrol queue. Edits with higher ORES destructive values can be held for longer, while edits with higher ORES constructive values can be autopatrolled. Aasim - Herrscher of Wikis ❄️ 02:12, 17 September 2023 (UTC)
It has been proposed numerous times on this wiki to install the PageNotice MediaWiki extension to allow a uniform header to be automatically displayed at the top of every page in the Draft namespace. See Wikipedia:Village_pump_(idea_lab)/Archive_39#Universal_notice_for_draftspace for what I understand to be the most recent discussion, but it was suggested as far back as 2014.
I'm pleased to report that the extension is now available for testing. The wiki used for testing PageNotice is Beta English Wiktionary (not Wikipedia).
I'm happy to grant established English Wikipedia users temporary admin rights on your Beta English Wiktionary account if you would like to test this new feature. There is no Draft namespace on Wiktionary, but you can test with any other namespace if you wish. Instructions for using the extension are at mw:Extension:PageNotice.
If there is a desire to enable the extension on this wiki (the non-beta English Wikipedia), community consensus would need to be demonstrated.
I'd prefer it to go above the page title, just like a banner would, rather than below.
When viewing or editing a page, it's hard to tell where the banner is coming from. Just like editors with the requisite permissions for editnotices on Wikipedia have the "page notice" and "group notice" links in the editor, there should be a link to the namespace banner MediaWiki page so that editors can edit it if desired.
@Sdkb re your second point, I feel this is something that should be handled locally, similar to the {{editnotice load}} system for editnotices.
The first point is a more interesting one. The placement of the notice at the top of the page content, beneath the title, makes it valuable for boilerplate text or disclaimers of the kind that this project might want to include at the top of Draft pages. What is the use case that you are imagining, where the notice goes above the page title? I note that in Vector 2022 skin, this would span the whole width of the page, deepening the disconnect between the page content (which is in a fixed-width panel) and the page notice. This, that and the other (talk) 03:16, 14 September 2023 (UTC)
The use case would be the draft namespace. Idk, it just seems more intuitive to have it above the title, and less like it could be confused with the actual page content. {{u|Sdkb}}talk03:19, 14 September 2023 (UTC)
Is there a technical reason why citations defined outside of sections can't be previewed, or would rendering them within the section preview be considered undesirable behavior? Photos of Japan (talk) 07:36, 17 September 2023 (UTC)
You'd have to do some fairly complicated stuff to get the definitions from outside the section in order to bring them in. The Phab task (T146072) User:Nardog added here describes one option of that "complicated stuff" for the context of potential section editing in VE; there are other ways it might be done that might work better for non-VE use cases. Anomie⚔13:21, 17 September 2023 (UTC)
Recently I have been seeing very small text size on some articles but not others, and on this very page, when using an iPad. I haven't figured out what it is about the page that changes the size. (Might affect tablet and not phone, would have to confirm on the latter.) When I say tiny, I mean so small that that Minerva is as hard to read as Vector and Vector 2022. Anyone know what changed? ⁓ Pelagic ( messages ) 18:16, 12 September 2023 (UTC)
Are you sure you didn't switch to the desktop site ? When you click the Aa in the addressbar on the device and click 100%, does that fix it ? —TheDJ (talk • contribs) 19:07, 12 September 2023 (UTC)Mobile web, 100% zoom
Yes, 100% sure, TheDJ. And zoom is 100%. I'll work on uploading some screenshots. Happens when logged-out also.
Since other people don't appear to be experiencing the same, I’ll test on some other devices. Maybe it’s “a me thing”. It looks fine on iPhone.) ⁓ Pelagic ( messages ) 21:25, 15 September 2023 (UTC)
On two old iPads that I keep around for testing it looks “normal”. 15–18 words per line, my top post line-breaks at: on/this, that/changes, latter.)/When, Vector/2022. (5 lines total) iOS 9.3.6 and 12.5
For comparison: desktop web using Vector 2022 with all sidebars (menu, ToC, tools) hidden, I get same line breaks on all three tablets: out/what, so/small. Portrait not landscape orientation in all cases. ⁓ Pelagic ( messages ) 22:39, 15 September 2023 (UTC)
Replicated on another iPad also running iOS 15. Confirmed so far on Safari 15.6.1, and Safari 15.6.6 on iOS 15.7.7. Viewport doesn't appear to be a factor: 810 x 977 and 768 x 921 affected; 768 x 922 and 928 unaffected.
Another relevant observation: I get a flash of normal larger text (FONLT?) and then the size changes to small.
1) The mobile web has a font size control in the settings page. Is that ticked?
2) On Special:Preferences you can disable responsive behaviour in the skin options. It applies to all skins. Has that been unchecked accidentally? Jdlrobson (talk) 16:16, 17 September 2023 (UTC)
Resolved
You are invited to join the discussion at Talk:Opinion polling for the 2023 Spanish general election § What happened to the References?. -- Marchjuly (talk) 22:29, 23 July 2023 (UTC)
On July 18, the Wikimedia Foundation launched a survey about the technical decision making process for people who do technical work that relies on software that is maintained by the Foundation or affiliates. If this applies to you, please take part in the survey. The survey will be open for three weeks, until August 7. You can find more information in the announcement e-mail on wikitech-l.
The new version of MediaWiki will be on test wikis and MediaWiki.org from 25 July. It will be on non-Wikipedia wikis and some Wikipedias from 26 July. It will be on all wikis from 27 July (calendar).
Moved to Wikipedia:Administrators' noticeboard § Massive wide-ranging IP block on Airtel India users
Works for me. Try this: at Preferences → Recent changes, unset "Group changes by page in recent changes and watchlist" and save; then set it again and save. --Redrose64 🌹 (talk) 21:21, 20 July 2023 (UTC)
Done — TheresNoTime (talk • they/them) 11:17, 31 July 2023 (UTC)
Moved from Wikipedia:Teahouse § Broken interwiki links
Resolved
Fixed (by abandoning that particular template entirely and switching to ordinary {{infobox settlement}}) * Pppery * it has begun... 17:56, 9 August 2023 (UTC)
02:08, 16 August 2023 (UTC)
As part of the changes for the better diff handling of paragraph splits, improved detection of splits is being rolled out. Over the last two weeks, we deployed this support to group0 and group1 wikis. This week it will be deployed to group2 wikis. [37]
Comment: I am unarchiving this stale discussion, due to finding another person who experienced a similar bug, @2NumForIce
)
Courtesy link: meta:Requests for comment/Allow thanking bots Edward-Woodrow :) [talk] 17:50, 4 September 2023 (UTC)
icon)
Note: I added the Phabricator ticket. I have no idea which tags to add. QuickQuokka [talk • contribs] 18:09, 9 September 2023 (UTC)
Courtesy ping: LittlePuppers QuickQuokka [talk • contribs] 18:37, 9 September 2023 (UTC)
199.208.172.35 (talk) 15:35, 12 September 2023 (UTC)
Mathglot (talk) 22:39, 10 September 2023 (UTC)
I'll pass this on! — TheresNoTime-WMF (talk • they/them) 09:48, 15 September 2023 (UTC)
![[2]](https://snipboard.io/VCFhRs.jpg){kind=link}
![[3]](https://snipboard.io/OWISc3.jpg){kind=link}
