Category:Wikipedians in Nizhny Novgorod is a redirect to Category:Wikipedians in Nizhny Novgorod Oblast but populated by Template:User Nizhny Novgorod in such a way that it's not possible to simply amend the user category. Instead it seems to be detecting the redirect and populating it by default. Can anyone untangle this one? Timrollpickering (Talk) 10:56, 6 May 2019 (UTC)

It can add a wanted category directly and call ((User Russian city)) with nocat = yes. PrimeHunter (talk) 11:15, 6 May 2019 (UTC)

Sorry I tried that but it doesn't seem to do the trick. Timrollpickering (Talk) 23:10, 6 May 2019 (UTC)

Eventually I reverted recent edits back until it got to the previous settings. This seemed to be the only solution. Timrollpickering (Talk) 23:15, 6 May 2019 (UTC)

You tried to add nocat = yes to ((userbox-2)) (which is only called for |oblast) instead of ((User Russian city)). I have done the latter now.[1] PrimeHunter (talk) 00:14, 7 May 2019 (UTC)

Can someone tell me the difference between these two redirects: Doña Ana County Jane Doe and Dona Ana County Jane Doe as one works (goes to the correct portion of the target article) and the other doesn't (goes to the top of the target article) Thanks GrahamHardy (talk) 07:14, 7 May 2019 (UTC)

@GrahamHardy: Doña Ana County Jane Doe redirects to List of unidentified decedents in the United States, which doesn't mention Doña Ana County. Dona Ana County Jane Doe redirects to a different article, List of unidentified murder victims in the United States, which does have a section about her. -- John of Reading (talk) 07:21, 7 May 2019 (UTC)

Doh! Thanks GrahamHardy (talk) 08:42, 7 May 2019 (UTC)

She was moved from the decedents list in January 2018.[2] The same user made the redirects in 2015 but forgot to update one of them. PrimeHunter (talk) 08:52, 7 May 2019 (UTC)

I have updated categories, but believe that Dona Ana County Jane Doe should be deleted as it is now identical to Doña Ana County Jane Doe, do people agree ? And what is the best mechanism for achieving this ? Thanks GrahamHardy (talk) 09:20, 7 May 2019 (UTC)

Further to the above I will remove all categories from Dona Ana County Jane Doe to stop duplication is categories. GrahamHardy (talk) 09:55, 7 May 2019 (UTC)

@GrahamHardy: Neither redirect was created recently (both were created in 2015: Doña; Dona), so should not be deleted without a formal WP:RFD. --Redrose64 🌹 (talk) 13:48, 7 May 2019 (UTC)

As one has an ñ and the other doesn't, they're not identical. The latter can be a redirect to diacritic. BlackcurrantTea (talk) 14:06, 7 May 2019 (UTC)

Is there any way to enable full-width/%-based width automatic resizing for images? I know Meta uses a ((Image)) template with custom .css to do that but we don't have that. Am I missing a more obvious solution that already exists? --qedk (t 桜 c) 15:53, 6 May 2019 (UTC)

QEDK yes, using template styles and the File syntax has a 'class' param that no one knows about ;) See how I used that on the header image of the Wikidata main page in d:Template:Main Page/Header new and d:Template:Main Page/Header new/styles.css. —TheDJ (talk • contribs) 07:12, 7 May 2019 (UTC)

Thanks a lot, that certainly helps! I asked the ((Image)) salting admin to unsalt it for this purpose, let's see. --qedk (t 桜 c) 17:53, 7 May 2019 (UTC)

see ((scalable image)) Frietjes (talk) 19:20, 7 May 2019 (UTC)

My watchlist is set for 30 days and 1,000 items. It's showing roughly 100 items starting at April 29, 2019, 9:50a (it's now only May 4). I want to see back to Apr. 20, and this happened once before (it was fixed somewhere behind the scenes, not by me), and apparently the starting date-time was a clue last time. Nick Levinson (talk) 20:44, 4 May 2019 (UTC)

Mine's showing the full list (250 items, which is my setting). Did you accidentally limit it to a single namespace or save a filter? Whatamidoing (WMF) (talk) 01:50, 5 May 2019 (UTC)

Does it say "Days to show in watchlist: 30" and "Maximum number of changes to show in watchlist: 1000" at Special:Preferences#mw-prefsection-watchlist? What does the first line of the "Watchlist options" box at Special:Watchlist say? For me it says "Below are the last 300 changes in the last 70 hours". There was once a problem with some integer days which could be fixed by setting a decimal like 29.9 instead of 30. PrimeHunter (talk) 09:52, 5 May 2019 (UTC)

Today, May 6, it's showing May 2-6 only and roughly 200 items only. Thus, the earliest date now follows the latest date, which is not the old problem but a different one. With all filters cleared (thus no active filters), the date range is the same but the item count is roughly 350 (including VP(T)). (I tried a quick and more precise count by searching for strings but got bizarre behavior: e.g., diffhist got some and (diff | hist) got others. No namespaces are checkmarked, so I should be seeing whatever is watchlisted in any namespace (I probably don't have anything in most namespaces). It does say "Days to show in watchlist:"/"30" and "Maximum number of changes to show in watchlist:"/"1000" at Special:Preferences#mw-prefsection-watchlist. I have no box titled Watchlist Options. I do remember that commonly there is a count of what's on the page, but I don't have that now. A browser search of the page for "the last" (which is part of the string you describe) finds nothing. I tried changing 30 days to 14 and back again on the watchlist page, but with no effect. Nick Levinson (talk) 01:43, 7 May 2019 (UTC)

The quoted messages were because I have enabled "Use non-JavaScript interface" at Special:Preferences#mw-prefsection-watchlist. Without that I see "300 changes, 2.9 days" in a box to the right. Have you tried a decimal like 29.9? PrimeHunter (talk) 09:13, 7 May 2019 (UTC)

Today, the range is May 2-7, I think the earliest date-time is the same now (time 18:22) as it was yesterday, though not what it was on Saturday, when it was an earlier date and, I think, earlier in the first day.

I tried 29.9 on the prefs page and then in the watchlist's URL but those made no difference. The latter change reflects above the watchlist items, where the menu title now says "1,000 changes, 29.9 days".

I cleared all filters again. This time, the earliest date-time, still May 2, is a little later for the time, suggesting that maybe the number that should be 1,000 is actually smaller but not saying so.

I removed VP(T) from my watchlist, since that's a frequently-edited page, so I hoped maybe earlier changes for other pages would show up. The odd result was that doing so extended the range a bit, May 1-7, but then I saw that the filters had come back (unsolicited by me), so I removed all of the filters, and the watchlist shrank to May 2-7. I don't understand how it shrank when filters went away.

I cut the intended number of changes from 1,000 to 100. That made the watchlist show 8 (eight) changes only, all for May 7. The URL says "limit=100".

I edited the URL to "limit=10000". That got me back to April 8. So now I have a kludge, but that's not the best solution.

I edited the URL from 29.9 to 30 and the start date became May 2 again, but then I saw that limit had become 1000, so I changed that to 10000 again with days left at 30, which got me back to April 7, a day earlier than a few minutes ago. A kludge, but not a happy kludge.

The basic problem is new, discovered last Saturday. My previous visit, on April 20, was uneventful, so the problem arose sometime between April 20 and May 4.

Nick Levinson (talk) 00:52, 8 May 2019 (UTC)

@J JMesserly, Hgrosser, and Zyxw: ((Birth-date)) and similar templates, for some reason, convert MDY dates to DMY dates, but only if no time zone is specified. The templates don't have |mf= or |df=.

• ((start-date|November 22, 1964)): November 22, 1964 (1964-11-22)
• ((start-date|21:00 EST, November 22, 1964)) 21:00 EST, November 22, 1964 (1964-11-23UTC02)

Does anyone know how those templates actually work? There's clearly something wrong with them, and I noticed this years ago but I don't think I did anything about it. Jc86035 (talk) 10:28, 8 May 2019 (UTC)

Fixed. The ((start-date)), ((birth-date)), ((end-date)), and ((death-date)) templates all now display dates as entered. Same examples in DMY format (with |tz=y added for timezone as per documentation):

• ((start-date|22 November 1964)) → 22 November 1964 (1964-11-22)
• ((start-date|21:00 EST, 22 November 1964|tz=y)) → 21:00 EST, 22 November 1964 (1964-11-23UTC02Z)

-- Zyxw (talk) 15:57, 8 May 2019 (UTC)

I'm getting a JavaScript error window popup when I preview a page: https://imgur.com/a/Kiv9oCJ RedWolf (talk) 18:50, 8 May 2019 (UTC)

@RedWolf: I get the same error, in my browser console. Does it go away when you disable Twinkle? It does for me. It may have been caused by Special:Diff/896151420 (pinging Amorymeltzer). Suffusion of Yellow (talk) 19:02, 8 May 2019 (UTC)

Indeed, I think it's the mw.config.get('wgCurRevisionId') !== mw.config.get('wgRevisionId') check in Mediawiki:Gadget-twinklefluff.js that's the culprit. wgRevisionId === 0 when edit or previewing, apparently, so Twinkle.fluff.oldid is getting called even when there's no mw-revision-info element. Suffusion of Yellow (talk) 19:16, 8 May 2019 (UTC)

Yup! Just pushed a quick fix; should take a few minutes to go through. Thanks for the ping, sorry for the inconvenience! ~ Amory (u • t • c) 19:19, 8 May 2019 (UTC)

That was quick! I no longer get the error, and the new "restore" button works for me. Suffusion of Yellow (talk) 19:27, 8 May 2019 (UTC)

• Yes, it's fixed. Thanks. RedWolf (talk) 19:40, 8 May 2019 (UTC)

The font size on this article from the discography onward is too small, but I'm unable to locate any unclosed font tags that are causing it. Can someone resolve this? Home Lander (talk) 02:48, 9 May 2019 (UTC)

 Fixed. The unclosed tag was in the preceding section. – Ammarpad (talk) 04:05, 9 May 2019 (UTC)

In the visual editor, if I pull up the citation tool, automatic tab, paste https://www.history.navy.mil/content/history/nhhc/research/histories/ship-histories/danfs/a/augustus-holly.html into the box, and click Generate, it changes the hostname in the URL. Any idea what's going on? -- RoySmith (talk) 16:03, 8 May 2019 (UTC)

I'm guessing it's because VE takes url from that tag:

<meta property="og:url" content="http://public1.nhhcaws.local/content/history/nhhc/research/histories/ship-histories/danfs/a/augustus-holly.html">. In code editor url stays the same (history.navy.mil). --MarMi wiki (talk) 16:22, 8 May 2019 (UTC)

MarMi wiki, the site publishes that as the ‘better url’ and also the one that should be used when sharing on facebook, twitter and google previews. That site really should be fixed. —TheDJ (talk • contribs) 06:20, 9 May 2019 (UTC)

Is there any reason why collapsible blocks (like Hidden, NavFrame, mw-collapsible) do not work in mobile version? I understand that this is because the code that deal with collapsible elements are in Common.js and not in Mobile.js, but is there any reason why this is a case? Screen space is more valuable on mobile devices, so mobile users would benefit from this feature even more than desktop users. Alexei Kopylov (talk) 22:22, 2 May 2019 (UTC)

@Alexei Kopylov: Will this be helpful? Adithyak1997 (talk) 03:11, 3 May 2019 (UTC)

No. I have not find an answer there why collapsible element are not collapsible in mobile version. Alexei Kopylov (talk) 14:56, 3 May 2019 (UTC)

@TheDJ:: do you know why mw-collapsible is supported by desktop version and not by mobile version? Alexei Kopylov (talk) 08:39, 9 May 2019 (UTC)

Alexei Kopylov, because no one put effort into making that so, and because a lot of the content would be inaccessible (due to tiny controls) because no one ever implemented mw-collapsible with mobile kept in mind (it literally predates mobile phones). —TheDJ (talk • contribs) 08:49, 9 May 2019 (UTC)

Also it has ALWAYS been the case that editors should NEVER expect collapsing to work, which is why it is not advisable to use in primary content. —TheDJ (talk • contribs) 08:50, 9 May 2019 (UTC)

Thanks. Alexei Kopylov (talk) 08:51, 9 May 2019 (UTC)

(reposted with some changes from Template talk:U.S. premium television services)

((North American TV)) appears three times in ((U.S. premium television services)): As a child of ((U.S. premium television services)), of ((U.S. premium television services (variety))), and of ((U.S. premium television services (PPV))) (the latter two are children of U.S. premium television services):

• U.S. premium television services
  • U.S. premium television services (variety)
    • North American TV
  • U.S. premium television services (PPV)
    • North American TV
  • North American TV

Removing it from U.S. premium television services leaves two copies; removing it from U.S. premium television services (variety) and U.S. premium television services (PPV) would leave those templates without it, should they be used outside of the U.S. premium television services template. What's the best way to clean this up? BlackcurrantTea (talk) 07:52, 9 May 2019 (UTC)

I will make it optional to add ((North American TV)) in ((U.S. premium television services (variety))) and ((U.S. premium television services (PPV))). PrimeHunter (talk) 09:41, 9 May 2019 (UTC)

Done.[3][4][5] ((U.S. premium television services)) was the only template which duplicated ((North American TV)). I don't know whether any articles have multiple navboxes which each add ((North American TV)). PrimeHunter (talk) 09:51, 9 May 2019 (UTC)

Thank you! I've yet to find any with the multiple navboxes, but if I do I'll try using the same syntax. BlackcurrantTea (talk) 10:26, 9 May 2019 (UTC)

Hi, If I lost my photo, and would like to log in to my Wiki account, I could just key in the scratch code into the code box? do I have to reset the 2 factor authentication all over again once I got my new phone? CASSIOPEIA^(talk) 12:12, 9 May 2019 (UTC)

@CASSIOPEIA: if you lost your 2FA device then you need to reset your 2FA as soon as possible! If you are logged in (and you are editing so it looks like you are) you can go to Special:Two-factor_authentication and DISABLE your 2FA using one of your one-time-use-only scratch codes. When you get a new phone you can enroll as a new 2FA setup. — xaosflux ^Talk 13:31, 9 May 2019 (UTC)

Xaosflux Thank you for the info. Another question, what should I do, if I am logged out and lost my 2FA device at the same time, how do I get myself log in prior disable my 2FA by using one of the given ten scratch codes? CASSIOPEIA^(talk) 13:42, 9 May 2019 (UTC)

@CASSIOPEIA: You can use the scratch codes (once per code) to log in as well. See Help:2FA for more details. So if you lose your phone and are not logged in, you will burn at least 2 scratch codes to get in, then disable your 2FA so you can reset it. — xaosflux ^Talk 13:51, 9 May 2019 (UTC)

Xaosflux OK, I see, good to know. Thank you. If I enroll a "new" 2FA setup, will I get a new 10 scratch codes? or user would have "only one time" 10 initial scratch codes and no more even after set up for new 2FA code? CASSIOPEIA^(talk) 14:09, 9 May 2019 (UTC)

@CASSIOPEIA: yes, it will be just like you are brand-new. All "old" TOTP client connections and old scratch codes will be voided, you will get a new TOTP initiation secret and 10 new scratch codes. — xaosflux ^Talk 14:43, 9 May 2019 (UTC)

Xaosflux OK thank you very much for answering my questions. Appreciate it. Cheers. CASSIOPEIA^(talk) 14:52, 9 May 2019 (UTC)

For background please see User talk:Ivanvector/Archive 12#Information about MediaWiki user and mw:Topic:Uzbq12lb6qltymd0.

A user on an IP range that is locally blocked with account creation disabled has created an account on mediawiki, but that account will not unify to enwiki and the user cannot edit here. Is this expected behaviour? A couple more specific questions:

1. If an accountcreator creates an account with the same name here, will the two accounts unify or can they be unified manually?
2. Same, but via WP:ACC

TIA Ivanvector (^Talk/_Edits) 15:03, 9 May 2019 (UTC)

Yes, block account creation prevents both local creations and autocreations. Otherwise vandals on blocked ranges would just sign up at a different wiki before immediately turning up here. That doesn't happen. I believe it's simply no longer possible to manually create an account when the global account exists. Two solutions: Sign up on another range, or lift the AC on the block. If you want a CU to follow that up, I'm sure that can be arranged. -- zzuuzz ^(talk) 15:40, 9 May 2019 (UTC)

The servers seem slow today; I thought that our edits get saved, and the changes get propagated to the SQUIDs. But my previous post seems to have gotten lost (it's the same question I am posting right now). --Ancheta Wis   (talk | contribs) 15:28, 9 May 2019 (UTC)

We've since changed from using Squid to using varnish. (Basically same thing, just made by different people). However, if you are logged in, you never use varnish for viewing site content, so varnish is probably unrelated to whatever issue you experienced. Bawolff (talk) 22:27, 9 May 2019 (UTC)

If I go to Interstate 10#New Mexico and click File:I-10 New Mexico 5.JPG (also reproduced at right), I get what looks like a normal en:wp page for a Commons image, but there's almost no metadata: below the image I see No higher resolution available and I-10_New_Mexico_5.JPG ‎(410 × 308 pixels, file size: 21 KB, MIME type: image/jpeg) as would be expected, then Open in Media Viewer and the normal "this image is on Commons" box the width of the page, and immediately below that is the file history. Initially I figured that the file had been vandalized, but when I view it on Commons, I see ((Information)) and ((PD-self)) looking entirely normal. I wondered if it had been cached after recent vandalism and before reversion, but the file was last edited in 2014. Can anyone guess why these aren't displaying when I view the file on en:wp? This appears to be a skin issue; I've viewed it logged-in (Monobook) and logged-out with IE, Firefox, and Chrome, and in every browser I saw a normal description page when logged out and a bad description page when logged in. Nyttend backup (talk) 20:36, 9 May 2019 (UTC)

PS, on this account, I've disabled Media Viewer. If you've not, please try opening the file in a new tab. Nyttend backup (talk) 20:38, 9 May 2019 (UTC)

To be clear, the bottom section called "Metadata" is not missing. It is the "Summary" and "Licensing" sections at commons:File:I-10 New Mexico 5.JPG which are not displayed at File:I-10 New Mexico 5.JPG. Those are the only sections made by the wikitext in the Commons page. All the other sections are made automatically by MediaWiki. The wikitext is also missing at other wikis, e.g. wiktionary:File:I-10 New Mexico 5.JPG. For comparison, File:Example.gif works and displays the wikitext at commons:File:Example.gif. The missing sections are sometimes but not always displayed for me when I log out and reload the page. PrimeHunter (talk) 22:34, 9 May 2019 (UTC)

If we compare with the metadata for similar locations, what jumps out for me is simply the difference between cameras: File:US_Route_54_in_New_Mexico.JPG --Ancheta Wis   (talk | contribs) 22:54, 9 May 2019 (UTC)

Previously after you clicked "Publish changes..." the cursor automatically selected the edit summary box so you could immediately type, but now it is not and you must also click in the box before you can type a summary. Please change back to the original functionality or let me know how/where to fix or further comment on this, thanks. Reywas92^Talk 06:55, 10 May 2019 (UTC)

I have opened a ticket. —TheDJ (talk • contribs) 08:57, 10 May 2019 (UTC)

Can anyone tell me where the text (highlighted in yellow) comes from? The box above is MediaWiki:Movepage-moved. Thanks — Martin (MSGJ · talk) 07:31, 10 May 2019 (UTC)

@MSGJ: It is from multiple messages: MediaWiki:Movepage-moved-redirect/MediaWiki:Movepage-moved-noredirect, MediaWiki:Wikibase-after-page-move, and MediaWiki:Movepage-page-exists. — JJMC89 (T·C) 07:52, 10 May 2019 (UTC)

For another time, WP:QQX has an API link with all messages. PrimeHunter (talk) 08:41, 10 May 2019 (UTC)

Thanks both, that's very helpful — Martin (MSGJ · talk) 10:02, 10 May 2019 (UTC)

When I recently moved an article to another name, I noticed that an archive page of the article's talkpage wasn't moved, despite me including the talkpage in the move. I had to do that move separately.

Aren't archive talkpages supposed to follow the talkpage when they are moved to another name?

HandsomeFella (talk) 18:14, 29 April 2019 (UTC)

@HandsomeFella: Admins and page movers have the move-subpages right, giving them an extra checkbox to ask for the subpages to be moved. The rest of us have to move them one at a time, or ask for help. -- John of Reading (talk) 18:29, 29 April 2019 (UTC)

Why is this separate? If a page move is valid, wouldn't you want associated subpages to go along for the ride? Seems to me that it would make more sense to give admins and page movers the option to leave subpages behind and make moving them the default. --Khajidha (talk) 18:03, 4 May 2019 (UTC)

I've thought this for years. The number of occasions when people want to move pages but not their subpages must be very small. I could see an RfC on this topic being of use.  — Scott • talk 12:53, 10 May 2019 (UTC)

Per phab:T76263, it seems there are issues with rate limits and vandals exploiting that. Another extremely dumb thing is that even for people with the move subpages right, it isn't checked by default.. Galobtter (pingó mió) 13:02, 10 May 2019 (UTC)

Tracked in Phabricator
Task T222953

I agree wholeheartedly about that checkbox and have just requested that it be enabled by default. That's been irritating me for years as well and I'm kind of surprised that I didn't request this sooner. Thanks for the idea.  — Scott • talk 13:26, 10 May 2019 (UTC)

Hi,

When I try to compare the two most recent revisions in the Methemoglobin article as of 16:02 GMT on 10 May 2019, I get the error message:

Internal error Jump to navigationJump to search [XNWSnwpAMFUAABhNTAMAAAAL] 2019-05-10 15:02:56: Fatal exception of type "Wikimedia\Assert\ParameterTypeException"

I believe this is the right place to report MediaWiki bugs? IAmNitpicking (talk) 15:04, 10 May 2019 (UTC)

diff in the page history. I get a similar error, also for the revision. diffonly works. The changes look simple. PrimeHunter (talk) 15:29, 10 May 2019 (UTC)

@IAmNitpicking: Hi, see mw:How to report a bug. This one is already reported under phab:T222857. --AKlapper (WMF) (talk) 17:21, 10 May 2019 (UTC)

Thank you. IAmNitpicking (talk) 17:31, 10 May 2019 (UTC)

I seem to have caused this, but don't know how to correct it. The transclusion would be from all the blue links on Template:Tony Hillerman .

Yesterday, I created an article with a typo: The Fly on the Wal I immediately realized my error and moved it to The Fly on the Wall without a redirect. That same typo was in Template:Tony Hillerman, and I corrected that yesterday. For some reason, the titles on Template:Tony Hillerman were not transcluding to "What links here" So, I created a redirect with the typo, pointing to the correctly named page. When I go to "What links here", I see the transclusion under the redirect name only. But that template is not actually on the Redirect. How can this be corrected? And please advise where I messed up, so I can avoid it in the future. Thanks. — Maile (talk) 19:38, 10 May 2019 (UTC)

Your description is a little unclear but I guess you mean: Why does Special:WhatLinksHere/The Fly on the Wall list many articles under "The Fly on the Wal (redirect page)" when the redirect hasn't been used by the articles since this edit to ((Tony Hillerman))? It's because automatic updates of link tables after a template edit can take a long time. The article pages are often updated faster without updating the link tables used by WhatLinksHere. Category pages can also take a long time to update after a template has changed categories and the list of categories on the article has been updated. The link tables for a single article are updated right away if you make a null edit or any other edit to the article, but there is no need for a fast update here. A purge of the article only updates the article page and not the link tables. Summary: Don't worry, it will eventually be fixed automatically. PrimeHunter (talk) 23:38, 10 May 2019 (UTC)

Yes. That is exactly what I meant. Thanks. — Maile (talk) 23:51, 10 May 2019 (UTC)

Resolved

 – Was user-specific .js file. — xaosflux ^Talk 02:29, 11 May 2019 (UTC)

When on a soft 404 article page for non-existent article "ONLY-A-TEST", I would expect that clicking either the "Create" tab, or the "Start the article" link would have the same behavior, namely, opening the "Creating ONLY-A-TEST" page. However, in one particular sequence of events, this is not the case. When using the "Create" tab, I'm seeing an Error: No such action page instead.

This is likely caused by a defective url generated by the Create tab, that places the query string parameter delimiters (? and &) in the wrong place. This only happens, afaict, when adding a nonexistent article name to the url of an existing page in the browser address bar.

What I'm seeing:

To reproduce:

1. go to any existing article, e.g. Foobar
2. in the browser address bar, replace 'Foobar' by 'ONLY-A-TEST' and hit Enter.
   • browser goes to soft 404 page 'Wikipedia does not have an article by this exact name...'
   • Note the urls underlying the 'Create' tab and the 'Start the article' link are not the same
3. Click the Start the article link
   • Expected behavior: browser opens the edit window of 'Creating ONLY-A-TEST'
   • Actual behavior: this works correctly. (but compare 5b)
4. browser back-arrow to 'Wikipedia does not have an article with this exact name' again.
5. Click the Create tab. (Note the reversed query string delimiters.)
   • Expected: browser opens edit window of 'Creating ONLY-A-TEST'
   • Actual: browser gives No such action page. (Not surprising, given the defective url.)

Why doesn't the Create tab generate the same url as the "Start the article" link? Mathglot (talk) 23:08, 10 May 2019 (UTC)

@Mathglot:  Works for me both the link and the create tab go to https://en.wikipedia.org/w/index.php?title=ONLY-A-TEST&action=edit. Also I'm not getting anything with the "redirect=no" part, where is that coming from for you? — xaosflux ^Talk 23:33, 10 May 2019 (UTC)

When you are creating the page, are you using the visual editor by chance? — xaosflux ^Talk 23:33, 10 May 2019 (UTC)

Also, I notice you have added special redirect javascript here: User:Mathglot/common.js - can you try taking that out? — xaosflux ^Talk 23:35, 10 May 2019 (UTC)

(edit conflict) @Xaosflux: No, the text editor. And using a PC, haven't tried mobile. Did you follow the reproduction steps? I wonder if this is browser-dependent somehow? This was Opera 60.0, I'll try a few others. Mathglot (talk) 23:39, 10 May 2019 (UTC)

Oh, shoot! I just added that js; I'm sure that must be the problem, I'm so sorry. That was related to another issue that was recently solved by some JS, which probably needs to be adjusted, to not cause this problem. I'll have to be more aware of knock-on effects of js changes. Thanks, and sorry for the false alarm. Mathglot (talk) 23:39, 10 May 2019 (UTC)

The problem is that User:Mathglot/common.js always adds ?redirect=no. If the url already has a query string after a ? then it should add &redirect=no instead. ? starts a query string while & separates parameters in a query string. PrimeHunter (talk) 23:49, 10 May 2019 (UTC)

@PrimeHunter:, Yes, I realized that as soon as xaosflux made his comment. See WP:US/R#Redirect page tab behavior. Mathglot (talk) 23:57, 10 May 2019 (UTC)

Hello, I'm from ckbwiki. I want to add undo and redo buttons to ours Wikipedia. Who can help me to do that? What should i do? Thank you, ئارام بکر (talk) 11:48, 19 April 2019 (UTC)

@ئارام بکر: ,WP:UNDO is built in functionality, you can see there is a link at this random ckdbwiki page I just loaded here. Can you describe this a bit more? — xaosflux ^Talk 13:00, 19 April 2019 (UTC)

@Xaosflux: No, I mean... add them such as button in Source edit, not history page. Just i want the source code. ئارام بکر (talk) 13:47, 19 April 2019 (UTC)

You can already do that with CTRL + Z and CTRL + Y. Amaury (talk | contribs) 14:18, 19 April 2019 (UTC)

"New wikitext mode" at Special:Preferences#mw-prefsection-betafeatures has such buttons (and other changes). It's also an option at ckb:Special:Preferences#mw-prefsection-betafeatures. PrimeHunter (talk) 15:00, 19 April 2019 (UTC)

@PrimeHunter: and @Amaury: thank you both, but i am a mobile user (not computer user) and i am use source editor (not visual editor). Mmm... i see the tool (on enwiki) in my preferences and active it. I think so it is an Mediawiki gadget. And i want to add that tool to ours Wikipedia. ئارام بکر (talk) 16:25, 19 April 2019 (UTC)

Gadgets are local user scripts.

User:PPelberg (WMF), I think that you might want to consider this for the toolbar improvements project. Without this available as an in-editor tool, then I don't know how else an editor on a smartphone would undo/redo a change inside the editor. Whatamidoing (WMF) (talk) 21:51, 22 April 2019 (UTC)

User:Whatamidoing_(WMF), thank you for bringing this to my attention. I agree with both you and ئارام بکر in thinking it ought to be clear in the mobile wikitext editor how to undo/redo edits/changes to the document. ئارام بکر, does this ticket task:T222316 accurately represent what you are requesting? As for prioritization, while we are planning to redesign the mobile editing toolbar this quarter (see task:T211255), that work – for now – is limited to VE. Which leaves me with a question, ئارام بکر what do you like most about editing using wikitext on mobile? It's what you're used to? It is difficult/not possible to make a certain kin of edit using the mobile Visual Editor? Something else? PPelberg (WMF) (talk) 22:38, 1 May 2019 (UTC)

@PPelberg (WMF): Thanks for your answer! When i joined to Wikipedia, i used wikitext on mobile and i'm happy with using it rather than VE. What about the result of this discussion? Can i see the buttons on ckbwiki in the future? Thanks! -- ئارام بکر (talk) 08:10, 8 May 2019 (UTC)

@PPelberg (WMF): I am waiting:) Thanks! --ئارام بکر (talk) 17:55, 9 May 2019 (UTC)

@ئارام بکر: Thanks for the nudge ^ _ ^ The decision to add a way to visually "undo" and "redo" changes within the wikitext editor on mobile is something we plan to revisit some time after July of this year. I suspect hearing this isn't as satisfying in this moment as hearing a definitive "yes" or "no," but I'd rather give you the most accurate answer I can. If you are interested in staying up-to-date on the discussion around this issue, you might find value in "subscribing" to this Phabricator ticket. If anything else comes up, please, leave me a message on my talk page (User talk: PPelberg (WMF)).

@PPelberg (WMF): Thank you for your answer! I subscribed. OK! I come back (maybe to your talk page) for the result in July. Please notify me if you opened this case again. Thank you all! --ئارام بکر (talk) 19:08, 10 May 2019 (UTC)

@ئارام بکر: awesome and will do :) Also, if you'd be interested in trying out the new features we're working on for the mobile VisualEditor over the next few months, please ping me. I know you mentioned wikitext was the editing interface you've come to prefer, but wanted you to at least be aware! PPelberg (WMF) (talk) 03:31, 11 May 2019 (UTC)

Resolved

Hi, when I look at the history of Talk:United Nations I see this edit as adding "+5,874‎" and this edit as adding "+5,834‎". These appear to be the total size of the page rather than the size of the changes. Does anyone know why this error occurs? Thanks, CMD (talk) 14:52, 11 May 2019 (UTC)

@Chipmunkdavis: there are some oversighted edits around that time that are spoiling the diff counter. — xaosflux ^Talk 15:02, 11 May 2019 (UTC)

The page was deleted and restored several times. One of these actions was a selective restore, and some edits which were selectively not restored (and so remained deleted) were also oversighted. --Redrose64 🌹 (talk) 15:56, 11 May 2019 (UTC)

Thanks xaosflux and Redrose64, appreciate you clearing that up! CMD (talk) 17:48, 11 May 2019 (UTC)

See e.g. Special:AbuseLog/23957087, Special:AbuseLog/23956301, Special:AbuseLog/23955139, Special:AbuseLog/23954164. I don't think it was intentional in every case.

The following steps seem to cause this:

1. Find a page with multiple sections, and switch to the mobile view.
2. Choose a section somewhere in the middle, and click the edit icon.
3. In you are not already in source editing mode, switch to that mode without changing anything.
4. Modify the content in some way.
5. Switch to visual editing mode.
6. Again, modify the content in some way.
7. Publish the page.

The expected behavior is that the one section you were editing is modified. The actual behavior is that the entire page is replaced with the section you were editing.

Can anyone else reproduce this, before I report it on phab? Suffusion of Yellow (talk) 20:46, 11 May 2019 (UTC)

Confirmed - using Chrome 74 on Windows 10, with the devtools "device toolbar". rchard2scout (talk) 21:23, 11 May 2019 (UTC)

I did some more testing in my sandbox, and reported it on phabricator: T223023. (Also, @Suffusion of Yellow:, forgot to ping you for my first comment, sorry) rchard2scout (talk) 21:57, 11 May 2019 (UTC)

Is there a way to count how many times one template was used in article? Is there something like "What links here" for the template but with use count per article. --MarMi wiki (talk) 18:10, 10 May 2019 (UTC)

Some progress:

Just discovered that you can use regex in search, is it possible to find two or more instances of template in one article?

Something like this, but for two or more Inflation hits per article:

hastemplate:Inflation AND insource:/(\{\{ *Inflation *.*[\}\|])/ --MarMi wiki (talk) 01:39, 12 May 2019 (UTC)

Resolved

I am having a few issues understanding taxobox (the article in question is Amphinotus nymphula):

1. The image in the infobox is not appearing, when I used the [[File:]] argument earlier (which was wrong), it previewed once and disappeared. It still isn't appearing.
2. The italic page name don't work a lot of the times (|name= was not defined obviously), why?
3. When I placed the status/status_system/status_ref parameters at the end (after binomial_authority), the graphical representation did not appear.

--qedk (t 桜 c) 10:12, 12 May 2019 (UTC)

1. Fixed.[6] There were some bad characters which prevented the image parameter name from being understood. It was fixed when I copy-pasted the line in Firefox so I guess Firefox automatically removes them when copying.
2. Please be specific with a link to an example you think should have italics according to Template:Taxobox#Bold/italic markup.
3. Maybe you had a similar problem with bad characters when status/status_system/status_ref parameters were at the end. It works when I move them to the end.

PrimeHunter (talk) 11:31, 12 May 2019 (UTC)

status_system had similar bad characters before [7] so it still failed when you first moved the parameters to the top in [8]. PrimeHunter (talk) 12:23, 12 May 2019 (UTC)

@QEDK and PrimeHunter: The "bad characters" were non-breaking spaces and this is another instance of this problem. In some browsers (I think Firefox is one), merely cutting the bad line and pasting it back in will convert all the NBSPs to normal spaces, which are valid. --Redrose64 🌹 (talk) 12:26, 12 May 2019 (UTC)

See this diff for the italics problem. I don't know what I changed but after this change, the title appeared as italics. I didn't realise there was non-breaking spaces, which is weird, because I even copy-pasted it off another of my working articles (Ellen Clark's crayfish) and I was equally surprised. Maybe the spaces are being converted to nbsps in Safari? --qedk (t 桜 c) 13:05, 12 May 2019 (UTC)

Other than iOS (mentioned in that archived thread that I linked earlier), I know of no browser which converts normal spaces to NBSPs. Experimenting with Firefox and Opera, I find that if you have a line of Wikicode that contains NBSPs, and cut or copy this using Opera, the NBSPs are preserved, and in Firefox they are converted to normal spaces. In either browser, pasting into the edit window will not affect the resultant type of space.

Cut or copy from	Paste to	Effect on NBSPs
Firefox	Firefox	Converted to normal spaces
Firefox	Opera	Converted to normal spaces
Opera	Firefox	Preserved
Opera	Opera	Preserved

I can't check Safari, since the most recent version available for Windows refuses to load Wikipedia pages - it 'can't establish a secure connection to the server "en.wikipedia.org"'. --Redrose64 🌹 (talk) 14:28, 12 May 2019 (UTC)

Your diff added italics in binomial = ''Amphinotus nymphula''. This caused italics in the page title. Template:Taxobox#Bold/italic markup says: "Italicization must be done manually in all parameters. If the entry for genus or species (with manually added italics) matches the page title then the name of the taxobox and the title of the page will be italicized." It appears this "genus or species" rule also applies for binomial. Template:Taxobox#Italic page titles says: "If the value of |genus=, |species=, or |binomial= exactly matches the title of the page, and |name= is unspecified, the taxobox and page title will be italicized automatically." binomial is included here but the rule apparently only applies when the parameter value has italics. PrimeHunter (talk) 17:09, 12 May 2019 (UTC)

Still no idea as to where the nbsps came from (given that I copy-pasted it from another of my articles with working parameters). I clarified the |binomial= documentation as it's not exactly accurate. Thanks a ton, @PrimeHunter and Redrose64:! --qedk (t 桜 c) 17:23, 12 May 2019 (UTC)

Hi, is there any way to use <timeline>...</timeline>, for ex, the way I use it is Template:F1Laps2019. How can I change, the "color" parameter and using a hex code (or similar)? For ex: #FFFFFF. Because some colors like sky blue doesn't work.

I have this two pages, [9] [10], if this can help me. So the point agaiin is that I want to change the color:yellow for a hex code that didn't limit the colors. ShaGuarF1 (talk) 19:58, 12 May 2019 (UTC)

@ShaGuarF1: Usually, if I want to use a complicated extension that I've not used before, I find an existing page that uses that extension and play withe the parameters (previewing always, and not saving) in order to work out what effect various changes might have. Apart from that, the general documentation is at mw:Extension:EasyTimeline, and the main parameters that accept a colour value are: mw:Extension:EasyTimeline/syntax#BackgroundColors; mw:Extension:EasyTimeline/syntax#Colors; and see the list of valid named colours at mw:Extension:EasyTimeline/syntax#Predefined colors. --Redrose64 🌹 (talk) 20:24, 12 May 2019 (UTC)

Yeah, I have used it but having a pattern to do that templates. But not in a "complex way". I edit while writing this: well, I don't remember if I visited that page before (I did a little investigation in the past), but I now already did it, that was not so hard to understand. Thank you for the comment. ShaGuarF1 (talk) 20:51, 12 May 2019 (UTC)

Can you tell me why ((maplink|frame=yes|zoom=6|id=6250|type=shape|text=Interactive map of Western Sahara)) will not work on Western Sahara? The entry at d:Q6250 has c:Data:Western Sahara.map associated with it. ―Justin (koavf)❤T☮C☺M☯ 06:42, 11 May 2019 (UTC)

Hm. Similar problem at Palmyra Atoll: c:Data:Palmyra Atoll.map and d:Q123076#P265 are centered properly but the interactive map here is in the middle of the Pacific... Why is this? ―Justin (koavf)❤T☮C☺M☯ 06:47, 11 May 2019 (UTC)

Koavf,Western Sahara fixed and reported as desirable improvement —TheDJ (talk • contribs) 20:54, 12 May 2019 (UTC)

Thanks! ―Justin (koavf)❤T☮C☺M☯ 20:54, 12 May 2019 (UTC)

@TheDJ: Can you figure out what's up with the Palmyra map? Why is it incorrectly centered? ―Justin (koavf)❤T☮C☺M☯ 20:55, 12 May 2019 (UTC)

Koavf, it's not. it was simply so far zoomed out that you couldn't see the atoll —TheDJ (talk • contribs) 21:10, 12 May 2019 (UTC)

Hm, when I scrolled in, I must have panned as well. Thanks. ―Justin (koavf)❤T☮C☺M☯ 21:14, 12 May 2019 (UTC)

00:48, 14 May 2019 (UTC)

For my last dozen or so edits I've been getting a captcha on every save, sometimes more than one, even though I'm not inserting any links. That wasn't happening earlier. Is something broken, is it a temporary setting because of a spambot running amuck somewhere, or what? Thanks. 67.164.113.165 (talk) 06:47, 13 May 2019 (UTC)

Please provide diffs of the edits where this occurred; or at the very least, the names of the pages concerned. It might be that there is an external link elsewhere in the same section. --Redrose64 🌹 (talk) 10:17, 13 May 2019 (UTC)

There are external links in the same section, but you're only supposed to get a captcha if you try to add a link that wasn't already there. This was at WP:AN. 67.164.113.165 (talk) 01:03, 14 May 2019 (UTC)

Is the problem still present? Did you get a captcha making this edit? Suffusion of Yellow (talk) 01:19, 14 May 2019 (UTC)

I don't think I had a problem with that one, but it is intermittent. I was able to save this current edit without it (i.e. I'll change this wording if I get a captcha). 67.164.113.165 (talk) 02:07, 14 May 2019 (UTC)

Did it happen today on WP:AN? Suffusion of Yellow (talk) 02:09, 14 May 2019 (UTC)

I made a post on AN that inserted 2 new links, and I think I got 3 captchas. I made a later post that didn't add any links and don't remember with certainty whether there was a captcha for that one. 67.164.113.165 (talk) 02:47, 14 May 2019 (UTC)

•  Confirmed. I just tried (as Suffusion of Yellow alt 5) to make an edit to WP:AN#Administrators preventing me from having RfC about an Administrative action containing NO added links, and got a captcha. But edits to other pages, also containing existing links, saved with no issue. Strange. I wonder what's special about that page. Suffusion of Yellow (talk) 02:53, 14 May 2019 (UTC)

  @Suffusion of Yellow: if you can actually make the edit past the captch, and let us know the Diff, we can examine it in more detail. — xaosflux ^Talk 03:37, 14 May 2019 (UTC)

  @Xaosflux: I got a captcha for Special:Diff/896996976. I think it's the IPV6 address in ((ipvandal)). Suffusion of Yellow (talk) 03:42, 14 May 2019 (UTC)

  @Suffusion of Yellow: http://<an ip address> is "an external link", so that certainly could do it. — xaosflux ^Talk 03:45, 14 May 2019 (UTC)

  @Xaosflux: Except the external link was already on the page in that edit. The problem, I've figure out, is that http://<an IPV6 adddress> is not a valid URL at all. See RFC 2732. If I format the URL correctly, as http://[<an IPV6 adddress>], the problem goes away. So it looks like a bug in MediaWiki and a problem with ((IPvandal)). Suffusion of Yellow (talk) 05:21, 14 May 2019 (UTC)

  This edit should fix ((IPvandal)) to generate a valid link. Anomie⚔ 13:04, 14 May 2019 (UTC)

  Wow, interesting. I have to sign off soon but I appreciate your looking into this. Thanks. 67.164.113.165 (talk) 03:49, 14 May 2019 (UTC)

I have a suggestion for a new tool. Short story: When you find an unsigned post like this, you activate the tool and it looks in the edit history to find the probable commenting user. After finding a possible match, it prompts the user for confirmation. We accept and it either tags it as ((unsigned|49.14.98.130|11:53, 11 May 2019‎)) or it properly formats the post by itself and maybe even sends a "please remember to sign your posts with four tildes" notice. The automatic signing bot doesn't always work and when I encounter these in the wild I have to do a bunch of copy/pasting and editing just to get someone's post set up right. Thanks! Cyphoidbomb (talk) 05:38, 12 May 2019 (UTC)

@Cyphoidbomb: User:Anomie/unsignedhelper does part of this. -- John of Reading (talk) 06:25, 12 May 2019 (UTC)

The "automatic signing bot" is SineBot (talk · contribs), which doesn't always work because it is down at the moment (I don't know whether Slakr (talk · contribs) is aware, but this mention should notify them). Also, ((unsigned)) is not the best one for IP addresses - for IPs, use ((subst:unsignedIP|49.14.98.130|11:53, 11 May 2019‎)); for logged-in users, use ((subst:unsigned|Example|11:53, 11 May 2019‎)) - please also note the presence of subst: in both cases. --Redrose64 🌹 (talk) 11:53, 12 May 2019 (UTC)

You can also use ((subst:xsign|11:53, 11 May 2019‎ 49.14.98.130)), which lets you copy and paste the time/IP string directly from the edit history, and automatically detects whether or not it is an IP address. --Ahecht (TALK
PAGE) 14:10, 14 May 2019 (UTC)

In List of shooting sports organizations we have markup like this:

;Indonesia:
* [[IPSC Indonesia]], international member of [[International Practical Shooting Confederation|IPSC]].
* [[Indonesian Target Shooting and Hunting Association]], international member of [[International Shooting Sport Federation|ISSF]].

;Iran:
* [[Shooting Federation Islamic Republic of Iran]], international member of [[International Shooting Sport Federation|ISSF]].

;Iraq:
* [[Iraqi Shooting Federation]], international member of [[International Shooting Sport Federation|ISSF]].

;Israel:
* [[Israel Shooting Federation]], international member of [[International Practical Shooting Confederation|IPSC]] and [[International Shooting Sport Federation|ISSF]].

;Japan:
* [[IPSC Japan]], international member of [[International Practical Shooting Confederation|IPSC]].
* [[National Rifle Association of Japan]], international member of [[International Shooting Sport Federation|ISSF]].
* [[Japan Clay Target Shooting Association]], international member of [[International Shooting Sport Federation|ISSF]].

Is that an acceptable description list or should it be regarded as a pseudo-heading? Sam Sailor 05:54, 13 May 2019 (UTC)

It appears there's not any technical problem with the heading itself, so the best place to discuss this is at the talkpage of the article or talkpage of Accessibility MOS: Wikipedia talk:Manual of Style/Accessibility. – Ammarpad (talk) 06:01, 13 May 2019 (UTC)

That copy is inappropriate for two reasons at least: It mixes actual list types, and WP:LISTGAP. If you want to have these be description lists, you would need to do something like:

;Indonesia
:* [[IPSC Indonesia]], international member of [[International Practical Shooting Confederation|IPSC]].
:* [[Indonesian Target Shooting and Hunting Association]], international member of [[International Shooting Sport Federation|ISSF]].

(Note I removed the colon directly after the term, instead placing it on the next line to make intent clear.)

The list gap piece is trivial when you've decided how to do your list; that's fixing it so that there are not spaces between the terms and descriptions:

;Indonesia
:* [[IPSC Indonesia]], international member of [[International Practical Shooting Confederation|IPSC]].
:* [[Indonesian Target Shooting and Hunting Association]], international member of [[International Shooting Sport Federation|ISSF]].
;Iran
:* [[Shooting Federation Islamic Republic of Iran]], international member of [[International Shooting Sport Federation|ISSF]].

I am on the fence about whether this is an appropriate kind of description list. I tend toward no, but you should invite others to discuss here (per WP:MULTI) from WP:Accessibility and/or WP:WikiProject Accessibility to see if they agree. --Izno (talk) 12:51, 13 May 2019 (UTC)

• This doesn't appear to be a technical issue, just an editorial one (how to style an article) - there are better venues than VPT (the article's talk page, a MOS page, etc) to hash this out. — xaosflux ^Talk 13:21, 13 May 2019 (UTC)

  Not really. There are two technical questions and the style question even though asked only as a single question; the former two I answered, but regardless, we should not split the discussion now that we're having it here (so it could be moved wholesale if interested). --Izno (talk) 13:57, 13 May 2019 (UTC)

  @Izno: can you clarify the technical issues, along the lines of "What were you expecting and what was the actual result" ? — xaosflux ^Talk 14:02, 13 May 2019 (UTC)

@Izno: It looks like you're still getting a listgap in there, apparently MediaWiki gets confused when the last colon line before a semicolon line has nested lists. It's easy enough to work around though by including an empty colon line:

;Indonesia
:* [[IPSC Indonesia]], international member of [[International Practical Shooting Confederation|IPSC]].
:* [[Indonesian Target Shooting and Hunting Association]], international member of [[International Shooting Sport Federation|ISSF]].
:
;Iran
:* [[Shooting Federation Islamic Republic of Iran]], international member of [[International Shooting Sport Federation|ISSF]].

If you want to visualize listgaps, you could use something like User:Anomie/listgap.css. Anomie⚔ 12:17, 14 May 2019 (UTC)

Interesting. I wonder if that's worth a phab task. --Izno (talk) 14:05, 14 May 2019 (UTC)

phab:T223277. --Izno (talk) 14:10, 14 May 2019 (UTC)

I'm sorry if this is the wrong place to ask, but I've so far failed to get help elsewhere. I'd really appreciate if someone who's good with templates could take a look at my proposed changes to ((lang-he-n)) (See Template talk:Lang-he-n § Idea: select version without niqqud) and give me some comments. Thank you in advance! Psiĥedelisto (talk) 03:51, 14 May 2019 (UTC)

Psiĥedelisto, I suggest raising that at Template talk:Lang, where at least two ((lang)) experts are active contributors. Justlettersandnumbers (talk) 20:12, 14 May 2019 (UTC)

I don't know what script or whatever determines whether a G13 is valid or not and gives a green OK if the last edit was more than six months before. So I don't know where to ask why it doesn't work in, for example, this case – it's not eligible, but the template reads "Missing parameter - Unable to calculate qualification to be deleted per WP:CSD#G13"? Can anyone shed any light? Justlettersandnumbers (talk) 18:34, 13 May 2019 (UTC)

I'd guess template talk:g13 to start. --Izno (talk) 18:58, 13 May 2019 (UTC)

@Justlettersandnumbers: It's the timestamp. Twinkle and YAAFCH take care of it, but this tagging was done manually, by just adding ((db-g13)) without a timestamp. The documentation for Template:Db-g13 should clear things up. ~ Amory (u • t • c) 19:46, 13 May 2019 (UTC)

Thanks, Izno, Amorymeltzer! Yes, I should have thought to look at the template documentation, but had not expected to find script documentation there too. So, does any of that work on mobile? We have a new(ish) user who seems to have chosen G13 as a career path, and I'd quite like to be able to guide him/her towards a better way of going about it – if there is one, that is. Justlettersandnumbers (talk) 20:06, 14 May 2019 (UTC)

Dunno what you mean about script documentation. I don't think either works well on mobile, they'd have to enter the ts manually I guess? ~ Amory (u • t • c) 20:52, 14 May 2019 (UTC)

It seems that anytime I use an indent/endash/bullet point, a space gets somehow added without me intending to. See 1 or 2 as an example. Using Firefox 66.0.2. YE ^{Pacific Hurricane} 23:48, 14 May 2019 (UTC)

@Yellow Evan: are you using WP:WikEd? Headbomb {t · c · p · b} 00:36, 15 May 2019 (UTC)

Yes. YE ^{Pacific Hurricane} 01:24, 15 May 2019 (UTC)

Hi all,

I've recently been working on changing the ((32TeamBracket-WSC2)) template from it's current form to one that uses named parameters to complete scores such as at 2019 World Snooker Championship. There's been a few WP:ACCESS issues, however I believe these to be at least in part fixed.

The issue I'm having, is that this template won't work for all championships. Whilst the number of frames for the final is somewhat static (it's been best of 35 frames since 1979); the sessions can have less/more frames than what is standard. It's supposed to be 8-9-8-10, however, some years certain sessions have overrun, and had different amounts (such as in 2016, where the third session only played 7 frames).

My question is, how would the best way to make a table such as the one that shows here as fluid like this. What my thoughts were, to have a parameter for each session, that said how many frames were supposed to be played in that session, and it would make the table fit those amounts of frames.

Please let me know if this makes any sense, as I realize it's a confusing question. Best Wishes, Lee Vilenski ^{(talk • contribs)} 08:14, 15 May 2019 (UTC)

Is there a way to show only a cropped portion of an image in an article?

In Derivative work, I have added a captioned, already-existing image, File:Marcel Duchamp, 1919, L.H.O.O.Q.jpg. For this article, I want to crop and zoom that image to eliminate the large whitespace border around the image, to ... make the Mona Lisa's mustache more obvious. The existing image is 800x1033 in size, and I'd want roughly the center 70% of the image, about 560x723. I was hoping there's an image template that lets me supply arguments for cropping, whether in pixel units or percentages.

The brute force method would be for me to download the image, crop it myself, and upload it under another name, duplicating the justification information from the original; but I was hoping there was a more elegant solution. I think a technical solution already exists at some level on Wikipedia, as when I mouse over the link derivative work, the image in the popup has clearly been cropped a bit around the edges (at least in my browser). On Sober Reflection (talk) 17:50, 15 May 2019 (UTC)

Yes; the ((Annotated image)) template can do it. The instructions are at Template:Annotated image#Cropping. In most circumstances, it's actually preferable to upload a separate image cropped as appropriate, as it makes it easier to translate the article into other languages which won't necessarily have a local equivalent of the template. ‑ Iridescent 17:57, 15 May 2019 (UTC)

@On Sober Reflection: ((CSS image crop)) also has this functionality. — xaosflux ^Talk 17:59, 15 May 2019 (UTC)

Also, if you are dealing with a free image, you can easily crop on Commons with commons:Commons:CropTool. — xaosflux ^Talk 18:01, 15 May 2019 (UTC)

CropTool overwrites the existing image, though; as I understand the question, it's "how do I display only a portion of the image without either uploading a second version, or permanently changing the existing image?". ‑ Iridescent 18:09, 15 May 2019 (UTC)

Iri It only overwrites if you tell it to, and most of the time you shouldn't per c:COM:OVERWRITE. GMG^talk 18:37, 15 May 2019 (UTC)

Yup, it is easy enough to make a new file, and it copies all the description/licensing for you as well (example here). — xaosflux ^Talk 18:54, 15 May 2019 (UTC)

Regarding ((CSS image crop)), you can try to do the reverse of this edit. But with a different pic, of course --Redrose64 🌹 (talk) 19:15, 15 May 2019 (UTC)

Naturally specification https://www.mediawiki.org/wiki/API:Login closely studied. I tried both Method 1 and Method 2.

In the first case, he writes: "Fetching a token via "action=login" is deprecated."

In the second case: "Invalid CSRF token".

How do I log in to the MediaWiki API? Are there examples? MaxiFix (talk) 10:17, 15 May 2019 (UTC)

@MaxiFix: You can use

api.php?action=query&meta=tokens&type=login&format=json

to get a login token, but as for actually using it and logging in; nobody knows :-) We're all plebeians here. If I figure it out, I'll let you know. Good luck. Guywan (talk) 15:39, 15 May 2019 (UTC)

MaxiFix, Method 1 only works with Special:BotPasswords. And you need to use a user-agent to keep track of the cookies. Tokens are available as described by Guywan. But what is it that you want to do ? Because the usecase should dictate if you use method 1, method 2 or an OAuth grant. —TheDJ (talk • contribs) 20:34, 15 May 2019 (UTC)

Moved from Wikipedia:Village pump (proposals) § Redirect page tab behavior

The "Article" tab on a redirect Talk page should go to the article, not to its redirect target.

The "Talk" and "Article" tabs should be a toggle: click one, and it should go to the other. Practically anywhere else in the encyclopedia, this is the case: if I'm on a Talk page of some article or project page, and click the "Article" or "Project page" tab, I go to the accompanying article, or Project page. Not so with redirect Talk pages, however. If I'm at Talk:This is a redirect page and I click the "Article" tab, it does not go back to the article, it goes to the target of the redirect. Not what I wanted. If I want the target page, I know how to get there.

I'd like to see this behavior changed. My guess is, a large majority of the time, someone on a Talk redirect page clicking the Article tab, wants to go back to the redirect article, not to the redirect target. If there's some legacy reason to keep this as the default behavior, then please at least offer an option in WP:Preferences to check a box assigning the "Article" tab (and "Project page" tab) to the article/project page, instead of to its redirected target page. Or, maybe some javascript whiz can cook something up for commons.js. Thanks, Mathglot (talk) 02:20, 9 May 2019 (UTC) ((ping)) please.

Hi, @Mathglot: I believe this belongs to where you first posted it; that is VPT, as it's more a technical decision than a social one. This is not something ideal for a preference setting but you can ask at Wikipedia:User scripts/Requests if someone is willing to write some script for this. The task phab:T5324 was asking the same thing and was declined because if this is implemented there's no way to go directly to the target page without manually editing the URL, among other reasons. – Ammarpad (talk) 04:34, 9 May 2019 (UTC)

Thanks! See WP:US/R#Redirect page tab behavior. Mathglot (talk) 06:23, 9 May 2019 (UTC)

@Mathglot and Ammarpad: I've reopened the Phabricator issue because I found the original reasons for the close didn't seem to directly address what changing this behavior would affect. For convienance, I'll include the main body of the comment I posted on the Phabricator ticket:

As far as I can tell, there is no advantage to immediately going to the target page after switching to the Page or Read tabs. One can only access these tabs from other tabs, and when I am in the Edit, Talk, or History tab, I am already examining the redirect in more detail, usually in preparation for editing it. In fact, I cannot think of a single instance where I wanted to go directly to the target page when I clicked on the Page or Read tab.

While this can easily be partly solved by the user script posted here, I still think it would beneficial if this was the core behavior. This would also make the interface more consistent; one is not taken to a redirect page's target after saving and editing, so I think the the Page and Read tabs should behave the same.

I don't know if reposting here is technically correct, since the VPT header says Bug reports and feature requests should be made in Phabricator, but I thought I would move it here from the proposals page so any potential further on-wiki discussion can be done here. e^πi (talk | contribs) 01:38, 10 May 2019 (UTC)

@Mathglot and Ammarpad: Fix ping. e^πi (talk | contribs) 01:40, 10 May 2019 (UTC)

According to Help:Redirect and my own experience, when a redirect is followed, a link to the redirect itself is just below the title. So it's one click away. DMacks (talk) 01:53, 10 May 2019 (UTC)

DMacks If it's scoped to the main article, that works fine, but if it's section scoped, it gets annoying (you may end up scrolling up the page significantly). So it's not always one click when going to the target page. But it's always one click away to the target page when one is at the redirect itself. e^πi (talk | contribs) 02:00, 10 May 2019 (UTC)

Good point. There have been proposals over the years to have the window decorations (header/tabs and left-column toolbox) frozen and only the article itself in a scrolling pane. That would resolve the section-link "I have to press the 'home' key or scroll up" situation. I think there is also a gadget that places a "home" link in each section-header line. DMacks (talk) 02:13, 10 May 2019 (UTC)

@DMacks: Also, I have a quick question: have you ever been at the Edit, Talk, or History tab of a redirect page and wanted to go to the redirect's target directly? e^πi (talk | contribs) 02:05, 10 May 2019 (UTC)

Sure, if I'm at a talkpage of an article that was merge/redirected while looking for previous discussion of a certain bit of article content, I then want to look at the current article content (follow redirect) or the history at the time of that discussion (History tab of the redirect), not at the redirect itself. If I'm editing a redirect, I sure want easy access to its target so I can work on extracting a self-contained section that has grown large enough to be a stand-alone. DMacks (talk) 02:13, 10 May 2019 (UTC)

I would agree that the intended behavior of the "Article" button, when on a talk/history page of a redirect, should be to link to the redirect itself. The number of times an editor wants to go directly from the redirect talk page to the redirect target is small compared to the number of times the editor wants to go to the redirect itself. This is not really a technical issue, as the technical implementation is straightforward if there is consensus to change the behavior, so I think VPP was probably a better venue. --Ahecht (TALK
PAGE) 22:02, 15 May 2019 (UTC)

Template:Infobox song contest entry has double bold in parameter "Name" in embded version. Template is protected and only template editors and administrators can edit it. Eurohunter (talk) 08:35, 17 May 2019 (UTC)

Template:Infobox song contest entry has no Name or name parameter. Click the "View source" tab to submit an edit request to a protected page. Clearly state which change you want. If you are thinking of the song parameter then the double bolding may be deliberate to indicate that the below headings are subheadings for song and not for the parent infobox. PrimeHunter (talk) 09:53, 17 May 2019 (UTC)

The Wikimedia Foundation is currently conducting a global consultation about communication. The goal is to bring Wikimedians and wiki-minded people together to improve tools for communication.

Phase 1 of the consultation is over – thank you to everyone who participated! – and we've published the Phase 1 report. The report summarizes what people have said and what we've learned, proposes a direction for the project, and asks specific questions to explore in Phase 2.

Very briefly, the proposed direction is that wikitext talk pages should be improved, and not replaced. We propose building a new design on top of talk pages that changes the page's default appearance, and offers key tools like replying, indenting and signing posts. To keep consistency with existing tools, the new design will be a default experience that existing users can opt out of. We also propose building features that experienced contributors want, including the ability to watchlist a single discussion, and the ability to move, archive and search for threads. Building these features may require some loss of flexibility, or small-to-medium changes in wikitext conventions. The goal is to only make changes that directly enable functionality that users really want.

You can see more information and discussion about the proposed direction in the Phase 1 report, including the results of new user tests and some of the quotations from Phase 1 discussions that led to this proposal.

Now it's time to start Phase 2!

We have six questions to discuss in Phase 2, asking for reactions to the proposed direction, and pros and cons for specific changes that we could make.

You can help by hosting a discussion at your wiki. Here's what to do:

1. First, sign up your group here.
2. Next, create a page (or a section on a Village pump, or an e-mail thread – whatever is natural for your group) to collect information from other people in your group.
3. Then start the conversation with the six questions listed in the Questions for Phase 2 section of the report.
4. When the conversation is concluded, the host should write a summary of the discussion on the Phase 2 community discussion summaries page, and report what you learned from your group. Please include links if the discussion is available to the public.

You can read more about the overall process on MediaWiki.org. If you have questions or ideas, you can leave feedback about the consultation process in the language you prefer.

Thank you! We're looking forward to talking with you. DannyH (WMF) (talk) 17:42, 17 May 2019 (UTC)

The Wikimedia Foundation has invited the various Wikimedia communities, including the English Wikipedia, to participate in a consultation on improving communication methods within the Wikimedia projects.

Phase 2 of the consultation has now begun; as such, a request for comment has been created at Wikipedia:Talk pages consultation 2019/Phase 2. All users are invited to express their views. Individual WikiProjects, user groups and other communities may also consider creating their own requests for comment; instructions are at mw:Talk pages consultation 2019/Participant group sign-up. (To keep discussion in one place, please don't reply to this comment.) Jc86035 (talk) 14:48, 18 May 2019 (UTC)

script execution order

One would expect imported scripts to execute in a linear fashion (or at least I did), but this does not appear to be so. How does one force the execution order of scripts imported into common.js (or other)? For reference, this is basically what I'm doing in my common.js:

   importScript("A");
   importScript("B");
   importScript("C");

Regards, Guywan (talk) 13:19, 16 May 2019 (UTC)

So, you could do this by using the ResourceLoader function of Mediawiki--which incidentally you should switch to anyway, because importScript is deprecated. mw.loader.getScript provides a means to load a script, and also provides a callback for when loading is done; you'd basically daisy-chain the callbacks. Like this:

mw.loader.getScript('/w/index.php?title=User:Foo/bar.js&action=raw&ctype=text/javascript').then( function () 
{
  mw.loader.getScript('/w/index.php?title=User:Foo/shoosh.js&action=raw&ctype=text/javascript').then(function ()
  {
    mw.loader.getScript('/w/index.php?title=User:Foo/pap.js&action=raw&ctype=text/javascript');
  });
});

If I may ask, though; why do you want them to run sequentially? Writ Keeper ⚇♔ 13:27, 16 May 2019 (UTC)

I don't, really. I was importing two scripts that added links to the Tools section of the sidebar, and sometimes the links would be in a different order. So, I was wondering if there was a way to prevent this. Thanks for the help! Guywan (talk) 17:57, 16 May 2019 (UTC)

@Writ Keeper: forgotten ping. Guywan (talk) 18:01, 16 May 2019 (UTC)

new lines

Also, completely unrelated: Is there a character code for inserting newlines into wiki pages? Regards, Guywan (talk) 13:19, 16 May 2019 (UTC)

@Guywan: it really depends on what you are trying to do, can you show an example of where you need a newline? The direct html code <br /> could be used in some cases, however we mostly try to avoid it in articles. — xaosflux ^Talk 14:34, 16 May 2019 (UTC)

Thanks,  Xaosflux:. Hopefully this might clear it up.

In the second example, what invisible character did the parser read that told it to insert a line-break, and how can I insert it programmatically? Regards, Guywan (talk) 17:57, 16 May 2019 (UTC)

Programmatically in what context? A JS user script? You can put a new line through the standard \n character within a string. For example: the call new mw.Api().newSection("User:Writ Keeper/sandbox","test","test\n\ntest"); led to this edit, newlines included. Does that help? Writ Keeper ⚇♔ 18:13, 16 May 2019 (UTC)

The parser wraps double \n\n characters as paragraphs. Single \n characters are included in the same paragraph tag. The "Magic" character is a \n.--Jorm (talk) 18:20, 16 May 2019 (UTC)

Thank you, @Writ Keeper and Jorm:. This helps very much. The problem was I only used a single \n, and not seeing what I wanted, assumed \n was stripped. Guywan (talk) 16:37, 18 May 2019 (UTC)

Resolved

I managed to mess up the AFD that I started here Wikipedia:Articles for deletion/List of tragedy films and TV programs. If someone can fix my error it will be most appreciated. MarnetteD|Talk 17:56, 18 May 2019 (UTC)

What's wrong with it? --Redrose64 🌹 (talk) 18:29, 18 May 2019 (UTC)

Nothing now. Thanks to QEDK for fixing it. MarnetteD|Talk 18:33, 18 May 2019 (UTC)

@MarnetteD: It was just an external link that you wikilinked accidentally, which broke the substitution, all fixed now! --qedk (t 桜 c) 18:47, 18 May 2019 (UTC)

D'oh! Thanks for letting me know QEDK. MarnetteD|Talk 18:51, 18 May 2019 (UTC)

That would just display brackets around the link: [this note]. The problem was mismatched brackets in [[https://en.wikipedia.org/w/index.php?title=List_of_tragedy_films_and_TV_programs&diff=next&oldid=895915235 this note]. When source code is displayed in the rendered page, always look for something mismatched. PrimeHunter (talk) 20:12, 18 May 2019 (UTC)

Wikipedia:Templates with red links lists all templates containing red links, but has not been updated in eight years. Can someone generate a new list? The existing subpages (Wikipedia:Templates with red links/001 through Wikipedia:Templates with red links/150) can be overwritten. Cheers! bd2412 T 23:09, 16 May 2019 (UTC)

• @Pppery: can do a database dump probably? -qedk (t 桜 c) 20:56, 18 May 2019 (UTC)

  @QEDK: What do I have do do with this? * Pppery * _survives 20:57, 18 May 2019 (UTC)

  @Pppery: I saw you do a query over at Quarry, so I thought you could help BD2412 with fetching this. That was quick! --qedk (t 桜 c) 20:59, 18 May 2019 (UTC)

In order to see my watchlist for a date range of about a week or two, I had to edit the number of items to 10,000, which I did in the watchlist URL by editing "limit=1000" to "limit=10000". Apparently, this has been disabled, with efforts to set the limit to 10,000 or 5,000 causing a redirect to 1,000, resulting in my seeing only about 50 items. What it shows me should be 30 days, and I've tried 29.9 and 14, but no matter, it's showing only May 14-18. The earliest on May varies, but sometimes it's an item from 7:37a. The first time it was around 7:37, then it was later in the day, and now it's 7:37a. I tried editing "days=30" in the URL to 40, but that auto-redirected to 30, with the earliest item now being 16:43p on May 14.

Removing VP(T), which has a lot of edits, from my watchlist did not solve the date range problem.

I'm not bothered by the closing off of one kludge, but we need something, so what solution or kludge can I try now?

Nick Levinson (talk) 19:46, 18 May 2019 (UTC)

The max is 1000, AFAIK, per the Prefs->Watchlist screen. When I go away for a week, I have to go through my Watchlist namespace by namespace. It's yet another way of making life a little more difficult for WP power users. T10681 is related, in that it also makes it harder to catch up after multiple days away. – Jonesey95 (talk) 20:30, 18 May 2019 (UTC)

What would be nice is if the watchlist URL accepted a parameter like &offset=20190512223000 in its query string, the same way that user contribs and page histories do; even better if it also allowed &dir=prev. That way, when you have finished going through your watchlist, you would note the date and time time of the most recent edit that you have examined, and on return from hols, set the offset to the date and time that you had reached, and work forward from that point. --Redrose64 🌹 (talk) 21:48, 18 May 2019 (UTC)

Support all of the above. At the very least let's get rid of the limit for extended confirmed users or for the non-Javascript watchlist. I'm fairly sure I used to be able to see more than 1,000 pages before the new watchlist was rolled out. Daß Wölf 04:12, 19 May 2019 (UTC)

@Daß Wölf: No, the "improved version of the Watchlist" moved from an opt-in beta feature to an opt-out live feature in July 2018; the 1000-change limit was introduced in August 2017. --Redrose64 🌹 (talk) 09:52, 19 May 2019 (UTC)

please fix Category treatment in watchlist (kludge is to exclude that)

I found a kludge, not my favorite and with some possibilities not tested: Leaving settings at 1,000 items and 30 days and with no active filters, open the Namespaces menu and select to exclude Category. You don't even have to exclude Category Talk.

If filtering is to include Category only, I get May 15-18, but if I want articles and Category only, then I get May 14-18, but if I want articles only, I get April 18 to May 18. That means that including Category alters the date range for one or more other namespaces. Most of my tests produced April 20-May 18. The difference between April 18 and April 20 may be the 1,000-item limit, so that's probably not a fault.

But Category should not be truncating my watchlist. If it must, then please add an explanation to the page so we'll know to select namespaces accordingly.

Nick Levinson (talk) 00:12, 19 May 2019 (UTC)

I reported this as a bug to Wikimedia's Phabricator. Perhaps the other suggestions in this topic should be reported there, too; I'll leave that to other people at their discretion. Nick Levinson (talk) 01:09, 19 May 2019 (UTC)

Are you seeing edits like this "(diff | hist ) . . Category:Heist films‎; 17:51 ‎Gjs238 ( talk | contribs‎ ) (Category:Japanese heist films added to category)"? When I'm watching a category and someone adds/removes multiple pages from it, they all show up in the watchlist, not just the most recent change (regardless of whether I'm watching the pages which were added/removed from the category), so when someone makes a real flurry of edits like that it can swamp the watchlist. Perhaps you've watchlisted some very active categories and those changes are not showing up but still counting towards the 1,000 limit somehow. Daß Wölf 04:33, 19 May 2019 (UTC)

Resolved

I've been getting this error intermittently for the last hour or two. It seems to happen on some pages a lot more than others. For instance, I got it once on VP(T) while leaving a reply in the section above, but after getting my last edit saved I can't access Virtual reality sickness at all anymore. Oddly this only seems to be a problem in one browser, despite the page showing a 503 code (server error). (I wasn't logged in on the other browser I tried.) Anyone else seeing this? Daß Wölf 04:21, 19 May 2019 (UTC)

• Not just you. --qedk (t 桜 c) 08:04, 19 May 2019 (UTC)
• It has been happening for days, on and off. - Sitush (talk) 08:06, 19 May 2019 (UTC)
  • I've seen it frequently in the last half hour, here and at Wikidata. The explanation on the error page about complicated templates doesn't fit the current error. If the revision history of this page loads several times without an error, and it hasn't changed, it shouldn't start producing error messages. Yet it did. BlackcurrantTea (talk) 08:29, 19 May 2019 (UTC)
• Can confirm, happened to me in id.wikipedia as well. Vulphere 10:00, 19 May 2019 (UTC)
• The issue is closed as resolved, if the issue persists (after the timestamp of the resolving comment only!), reopen the Phab ticket. --qedk (t 桜 c) 13:37, 19 May 2019 (UTC)

About a week ago, my computer started locking on all-capital letters when I capitalize a letter, but only when I'm typing in an English Wikipedia edit box--not in the edit summary box, and not in French Wikipedia or Wiktionary or non-Wiki sites. How do I turn off the automatic caps lock? (I'm using an iPad.) Loraof (talk) 15:41, 18 May 2019 (UTC)

This is a known issue, of which the cause has not been fully tracked yet. —TheDJ (talk • contribs) 16:25, 19 May 2019 (UTC)

13:03, 20 May 2019 (UTC)

Hello everyone, I'm from ckbwiki. I added this mediawiki page, but it doesn't work properly! Now what should i do to create ((SHORTDESC:description)) (The magic word) on our Wikipedia? Thanks! --ئارام بکر (talk) 17:26, 19 May 2019 (UTC)

@ئارام بکر: see phab:T184000 and related tasks. ckbwiki does not appear to have the shotdesc parser hook enabled (w:ckb:تایبەت:وەشان). — xaosflux ^Talk 17:36, 19 May 2019 (UTC)

@Xaosflux: but why? English wikipedia how did that? Why other wikipedias can't? I want to add the magic word in same way that enwiki did that. Any way? --ئارام بکر (talk) 19:04, 19 May 2019 (UTC)

@ئارام بکر: Using and changing the appropriate letter-code descriptions on Wikidata will do the same thing. Your wiki probably does not need the short description magic word. --Izno (talk) 19:46, 19 May 2019 (UTC)

@Izno: I know, but that gadget is more useful and easy than we go from ckbwiki project to wikidata and repeat again and again always. --ئارام بکر (talk) 23:04, 19 May 2019 (UTC)

Unless I'm missing something, that gadget makes a LOCAL description, it doesn't update the normally used wikidata entry. — xaosflux ^Talk 23:42, 19 May 2019 (UTC)

@ئارام بکر: I'll look into making the gadget work on other wikis (so you can edit the wikidata description without going to wikidata) while working on User talk:Galobtter/Shortdesc helper#Export to wikidata?; should be easy enough. Galobtter (pingó mió) 16:17, 20 May 2019 (UTC)

The code for Wikipedia:Database reports/Indefinitely blocked IPs has a size of 1,784,147 bytes, and as it's basically a big chart with lots of text, I expect that the complete page size isn't significantly larger than that. (It has no images at all, except the little up-and-down arrows, and images present on all pages, like File:Wiki.png and File:Wikimedia-button-for-homepage.png.) Conversely, http://proceedings.esri.com/library/userconf/fed16/papers/fed_86.pdf has a size of 27,717,239 bytes. Can anyone guess why the first page takes a lot longer to load than the second page? I just discovered the second page a few minutes ago; it's not as if I had it cached. And while Esri is an important player in its field (software for Web-mounted GIS), I suspect that their servers' speed and capacity are significantly less than Wikipedia's, with Alexa rankings of #7,072 and #5 respectively. And it's not just my computer; go to Wikipedia:Village pump (proposals)/Archive 158 and look for "Indefinitely blocked IPs" to see Galobtter's comment. Nyttend (talk) 21:06, 19 May 2019 (UTC)

PS, please note that Esri's self-hosted GIS sites tend to be mounted on arcgis.com, so I don't think the nature of the site would require them to have significantly faster servers. Nyttend (talk) 21:16, 19 May 2019 (UTC)

@Nyttend: Does the page still take a long time to load while you are logged-out? --AntiCompositeNumber (talk) 01:28, 20 May 2019 (UTC)

Interesting: it loads a lot faster, even though I have no scripts installed except one to hide a link in a certain infobox and another to give me the traditional "You have new messages" bar. Could it be the skin? I use Monobook. Nyttend (talk) 01:39, 20 May 2019 (UTC)

Nyttend, when you are logged in, you get a fully dynamically generated page, when you are logged out, you get a partially cached pre-rendered page which is slightly faster, but still not as fast as a fully pregenerated thing like a PDF. —TheDJ (talk • contribs) 10:53, 20 May 2019 (UTC)

Nyttend, they don't really compare. One is a website, the other a pdf. One is generated on the fly based on wikicode, the other is pre generated. —TheDJ (talk • contribs) 10:38, 20 May 2019 (UTC)

Hm, okay; I didn't realize that there was a difference from the server's perspective. I figured it sent me a bunch of 1s and 0s (about 14¼ million for the webpage and 221.7 million for the PDF), and once they'd all downloaded, the page was ready. Nyttend (talk) 11:03, 20 May 2019 (UTC)

Nyttend, yeah it's not like that. By that definition buying a factory printed book in the store vs having the store employee type over a book using a typewriter and handing you that copy would take the same time ;) —TheDJ (talk • contribs) 09:43, 21 May 2019 (UTC)

• Both of them took negligible time for me, which means, it's a problem with your internet connection, browser, intermittent server-side problems, OS, or a combination of them (not an exhaustive list). --qedk (t 桜 c) 18:09, 20 May 2019 (UTC)

I am having issues accessing https://tools.wmflabs.org/copypatrol/en, getting an error "HTTP error 500" message. Some of the other pages on Labs are not working, for example these ones are not working:

• https://tools.wmflabs.org/admin/tools
• https://tools.wmflabs.org/xtools/rangecontribs/index.php?
• https://tools.wmflabs.org/rangecontrib/
• https://tools.wmflabs.org/xtools/autoblock/index.php

Any assistance in getting this fixed would be appreciated. Thank you, — Diannaa 🍁 (talk) 12:41, 21 May 2019 (UTC)

Update: Some of these are now working, including CopyPatrol, but others are not. — Diannaa 🍁 (talk) 14:24, 21 May 2019 (UTC)

CopyPatrol has been fixed. The outage was due to phab:T223992. I assume it was the same issue that affected toolforge:admin/tools. @Diannaa: You should not be using any of those legacy XTools links. Range contributions is now part of core, just enter the IP range at Special:Contributions. I don't know anything about toolforge:rangecontrib but I assume it did the same thing. Autoblocks are viewable at Special:AutoBlockList. If you need to search by autoblock ID, currently you need to do this at Special:BlockList, putting #123 as the "IP address or username", where 123 is the autoblock ID. — MusikAnimal ^talk 14:29, 21 May 2019 (UTC)

Thank you, I will remove the old tools from my collection. — Diannaa 🍁 (talk) 14:34, 21 May 2019 (UTC)

Resolved

Template:Parenthetical referencing editnotice is supposed to place an editnotice that is only visible when someone opens up the editing window on an article but instead it places a BIG BRIGHT notice onto the article itself when it is in the Read-only mode. (See this lovely example at Hyde Park Picture House.)
If this editnotice worked the way it says it's supposed to work, it should only be visible to editors when they are editing a Harvard-cite'd article that the template has previously been placed upon. Someone, please fix it...I don't know how and have just discovered this editnotice and it would be really really REALLY useful to (hopefully) forestall editors not following an established Harv-cite referencing style on an article. Thanks, Shearonink (talk) 05:40, 21 May 2019 (UTC)

@Shearonink: Edit notices are created in a pseudo-namespace at Template:Editnotices/Page/___. I've created Template:Editnotices/Page/Hyde Park Picture House for you --DannyS712 (talk) 05:46, 21 May 2019 (UTC)

Thank you...but that sort of doesn't answer my question... There then needs to be additional documentation/instructions on the actual Template page to tell editors how to implement this template. And, does that mean that for every time I want to use this Template that I have to create a pseudo-namespace for it? Can't it be transcluded or something? On almost every single article that uses Harvard cites that I come across there is a continuing problem of editors adding the more-common straight inline citations and then that subsequently creates a funfunfun mess to clean up afterwards. Shearonink (talk) 05:58, 21 May 2019 (UTC)

@Shearonink: If you want to post a list at my talk page, I'd be happy to create edit notices for you --DannyS712 (talk) 06:00, 21 May 2019 (UTC)

Heh, you have no idea how long that list can get...I mean, now that I understand what to do I can do it myself but seriously...this is the only way to do it? Also, just to make sure I understand how to implement it... When the pseudo-namespace thingy is created it then will just appear on the article whenever the article is edited, it doesn't have to be added to the article itself as another bit of code? Thanks, Shearonink (talk) 06:09, 21 May 2019 (UTC)

@Shearonink: Correct --DannyS712 (talk) 06:25, 21 May 2019 (UTC)

Ah...ok, thanks. Since you understand how editnotices work would it be possible for you to add some instructions on the template itself? Editors not understanding how editnotices work seems to be a longstanding issue from the Template's talkpage... Thanks again, Shearonink (talk) 06:39, 21 May 2019 (UTC)

Shearonink, you may be thinking of ((Use Harvard referencing)), which is meant to be placed at the top of articles. – Jonesey95 (talk) 06:43, 21 May 2019 (UTC)

I might have been but I wasn't. I had already found that Template, used it and found it wanting for my purposes. I wanted and want an edit notice to boldly appear when the editing window is opened. In my opinion the "Use Harvard referencing" is a little too demure for most editors to really see it... Shearonink (talk) 07:22, 21 May 2019 (UTC)

Shearonink, well.. the template does say: "To create a new editnotice, please follow the instructions at WP:Editnotice" and that should explain everything right ? —TheDJ (talk • contribs) 09:40, 21 May 2019 (UTC)

Does it? I don't see that line of text anywhere on Template:Parenthetical referencing editnotice so I obviously missed that part when I tried to use the template and that's why I asked here. And if someone who is somewhat-experienced around WP misses it so might others. So far as I can tell the template at present reads as follows:

• To display this message when an editor edits an article, create an editnotice for the article containing ((Parenthetical referencing editnotice)). This template is not for use in articles. Editnotices are separate pages that display as message boxes shown about the edit window when a user edits a page.

Btw the present instructions at the How-to guide WP:Editnotice?... Well, DannyS712's explanation above was very clear, very concise, and I understood it right away. The "official" how-to guide instructions?...well, I had to ask here didn't I. Shearonink (talk) 17:00, 21 May 2019 (UTC)

@Shearonink: The methods for adding editnotices to specific pages differ according to whether or not you have the templateeditor user right or not.

If you do not have this right, you need to make an editnotice request by following the directions at WP:Editnotice#How to request an editnotice.

If you do have this right, you will find when editing any page (or section of a page) that one or two links are displayed at upper right, titled "Group notice" and "Page notice", which may each be red or blue. The "Group notice" one may be absent; but regardless, the one to click and edit is "Page notice". For example, when editing Actuary, there is only a "Page notice" link, which is blue, and goes to Template:Editnotices/Page/Actuary. --Redrose64 🌹 (talk) 19:41, 21 May 2019 (UTC)

@Redrose64 and Shearonink: also, User:DannyS712/Edit notice.js adds a link to a page's edit notice, without needing to edit the page --DannyS712 (talk) 20:05, 21 May 2019 (UTC)

All of my talk page below the edits of February 18, 2018 (15 months ago) were rendered illegible by this edit by Wikiacc. Why? How can it be fixed? Michael Hardy (talk) 19:45, 21 May 2019 (UTC)

... and now it appears I've fixed the problem, except that the thing that appeared to be causing it has been there for a long time, so it's a mystery why it suddenly happened now. Michael Hardy (talk) 19:57, 21 May 2019 (UTC)

An edit in February last year added a <math> tag without the balancing closing tag. Wikiacc's edit added three (balanced) pairs of <math>...</math>, the first of which was taken as the closing tag for the one added in February. --Redrose64 🌹 (talk) 20:16, 21 May 2019 (UTC)

Hello everyone, I'm from ckbwiki. I thought enwiki add this (and a sipmle) gadget like ckbwiki and arwiki also. If this community accept the gadget, please an interface editor create the gadget by following these instructions:

Create Mediawiki:Gadget-RedirectLinks with the following:

Display links to [[Help:Redirect|redirect]] pages in near-green cola

Then, create Mediawiki:Gadget-RedirectLinks.css with the following

/**
 * Display links to redirect pages in near-green cola
 * @revision 1.0 (2019-05-21)
 * @author ئارام بکر
 */
a.mw-redirect { color: #4e6f1c; }

Then, at MediaWiki:Gadgets-definition in appearance section add the following:

* RedirectLinks[ResourceLoader]|RedirectLinks.css

Finally, check the gadget and enjoy!

Goal: See all redirect page links with the different color ("near-green cola" for now).

Thanks:) --ئارام بکر (talk) 23:07, 21 May 2019 (UTC)

@ئارام بکر: I suggest you start by listing this at Wikipedia:User scripts/List and see if people start using it first. — xaosflux ^Talk 01:25, 22 May 2019 (UTC)

There is also User:BrandonXLF/GreenRedirects, which does something similar --DannyS712 (talk) 02:18, 22 May 2019 (UTC)

@Xaosflux: and @DannyS712: I do apology to User:BrandonXLF. I didn't know that this gadget (or script) is available already. However, thank you both! --ئارام بکر (talk) 15:41, 22 May 2019 (UTC)

Resolved

the obnoxious weird box being visible over pages that have a red linked category? See this image. ∯WBG^converse 12:14, 22 May 2019 (UTC)

See Template_talk:Sockpuppet#Sockpuppet_categories. According to the documentation, it's garish. -- zzuuzz ^(talk) 12:28, 22 May 2019 (UTC)

For the technical side, the image is from User:PedroLucas151204 with source ((sock|João Victor23567|confirmed)). It uses ((Sockpuppet/categorise)) which uses mw:Help:Extension:ParserFunctions##ifexist to test whether the corresponding category exists. PrimeHunter (talk) 12:37, 22 May 2019 (UTC)

The helper script is creating new cats and I don't see any reason for the message to exist. Inclined to revert. ∯WBG^converse 12:41, 22 May 2019 (UTC)

Just someone's way of throwing errors, it's not that garish imo, either way,  Fixed. --qedk (t 桜 c) 12:41, 22 May 2019 (UTC)

@Winged Blades of Godric: Please don't upload Wikipedia screenshots to imgur. My browser takes several minutes to display anything, and then it's the cookie information box which I am obliged to "accept" before I can see anything useful. Even when I have exited the website, my computer slows to a crawl necessitating at the very least close and restart the browser. In the past, my antivirus software has warned me about imgur and I don't want to reboot every time.

Instead, we have guidance at WP:WPSHOT. --Redrose64 🌹 (talk) 13:20, 22 May 2019 (UTC)

If something appears when you log out and it doesn't look like a browser issue (layout often is) then you can usually just link the page and describe it. If it's text then quote a sample. PrimeHunter (talk) 13:30, 22 May 2019 (UTC)

(edit conflict) Same with me, but it eventually opens and doesn't affect my system. Idk if this is helpful, but I upload stuff to imgBB (which downsizes it) and generally distribute the hotlink to the static image page. --qedk (t 桜 c) 13:43, 22 May 2019 (UTC)

I don't personally care as much about imgur screenshots as I do about un-search-friendly, unhelpful section headings like "What's", made out of the first word of the comment. But that's probably just me... -- Begoon 13:39, 22 May 2019 (UTC)

Begoon, Apologies. I ought to keep that in mind. ∯WBG^converse 15:44, 22 May 2019 (UTC)

Don't lose any sleep over it - just a peculiar niggle of mine. People in help forums I frequent do it a lot and it inexplicably irritates me whereas I don't usually care about small stuff like that. Illogical, but true... -- Begoon 15:53, 22 May 2019 (UTC)

I'm certain I checked the box when for some reason I wasn't signed in earlier this week. I turned off the computer on Monday and turned it on this morning.— Vchimpanzee • talk • contributions • 15:13, 22 May 2019 (UTC)

Your browser probably just cleared your cookies. Happens occasionally. --qedk (t 桜 c) 17:47, 22 May 2019 (UTC)

I have noticed a lack of purple links. But the last time I restarted the computer because of updates was Sunday.— Vchimpanzee • talk • contributions • 18:38, 22 May 2019 (UTC)

That's a definite sign your browser is clearing your history. Updates sometimes have something to do with it and sometimes don't. Cookies are set to expire after a while and while the Wikipedia "Remember me" cookie is indeed set for 365 days, it probably won't last that long anyway. --qedk (t 桜 c) 18:58, 22 May 2019 (UTC)

Resolved

 – Nothing was technically wrong here, but there is a suspect compromised account being dealt with and cleaned up after. See also Wikipedia:Administrators'_noticeboard/Incidents#Probable_compromised_admin_account. — xaosflux ^Talk 20:17, 22 May 2019 (UTC)

Sorry if this is off base, but I just came across User:NVPA8200. They have no contributions anywhere, and were created 16 hours ago. However, they are in the following user groups: edit filter manager, edit filter helper, account creator, autopatrolled, confirmed, event coordinator, extended confirmed, page mover, file mover, IP block exempt, mass message senders, new page reviewers, pending changes reviewers, rollbackers, and template editors. That being said, I can't find any log of the account being created or given these rights (see enwiki log and meta log). Any ideas? --DannyS712 (talk) 19:00, 22 May 2019 (UTC)

The logs are redacted. Galobtter (pingó mió) 19:08, 22 May 2019 (UTC)

@Galobtter: Oh, never mind then --DannyS712 (talk) 19:09, 22 May 2019 (UTC)

Yeah, it's not the first time, I think. Some people prefer CLEANSTARTs like this (not assuming anything about this account). --qedk (t 桜 c) 19:13, 22 May 2019 (UTC)

These logs probably shouldn't be revdel'ed, per Wikipedia:Revision_deletion#Log_redaction. I'm not sure if I can say who RD'ed it (erring on the side of caution). SQL^{Query me!} 19:21, 22 May 2019 (UTC)

• Per WP:ADMINACCT I'd like to know the reasoning behind this and why it's not visible with a reasoning for it. Clean starts don't mean you get logs redacted and all the perms...Praxidicae (talk) 19:24, 22 May 2019 (UTC)

• What Praxidicae said. Clean starts mean you create a fresh account and restart from scratch, they don't mean you get to keep those parts of your previous history you feel like keeping while jettisoning those parts you consider inconvenient. ‑ Iridescent 19:27, 22 May 2019 (UTC)

This also brings up a potentially much larger issue but I'll wait for the answer to the "why" before I get into it...also how can we apply WP:ADMINACCT to something like this where such an action is not visible to the larger community? I would think that any type of RD or even OS of this nature would warrant an explanation that would make it obvious to at least someone with a mop that there was good reasoning? Praxidicae (talk) 19:29, 22 May 2019 (UTC)

• Agree with Iridescent. I think the number of rights is primarily why they did not want a blanket new account — gaining this many rights takes a significant amount of limbs into each field, and that's a lot of time and effort. Although, and I am playing the devil's advocate here, what's the point of the EFM+EFH combination? --qedk (t 桜 c) 19:35, 22 May 2019 (UTC)

• @QEDK, the EFM+EFH combination is because they've literally granted the new account every single bit that it's technically possible for an admin to grant unilaterally. ‑ Iridescent 19:37, 22 May 2019 (UTC)

@Iridescent: The EFM bit worries me but I explained a little more below.

• I think I've put 2 + 2 together here so forgive me if I'm wrong and as I'm not a mop holder and have no ability to review the logs, this is merely based on what is obvious but I think an explanation is warranted given the admin appears to have created a second account, given it all the perms that aren't crat given and then redacted the logs. I apologize if this isn't allowed but they also don't have email enabled so I can't even ask privately. Would pinging the user here be any violation? I'm quite concerned given the recent compromises and the fact that there's no email but they've given themselves EFM access on a redacted account. Praxidicae (talk) 19:40, 22 May 2019 (UTC)

  It's kind of stupid to redact one log and leave the other one as-is. --qedk (t 桜 c) 19:44, 22 May 2019 (UTC)

  @QEDK: I think they were just unredacted by TonyBallioni --DannyS712 (talk) 19:47, 22 May 2019 (UTC)

  @DannyS712: There was a 2:00 Special:Log/rights log by Nv8200pa which was not redacted, immediately followed by a 2:09 redacted (now unredacted by Tony) log — that's why I called it stupid, since they redacted only one of them. --qedk (t 桜 c) 19:53, 22 May 2019 (UTC)

  @QEDK: One of them was for the hidden account, and one was for themselves, but I see what you mean --DannyS712 (talk) 19:54, 22 May 2019 (UTC)

• User:Nv8200pa is an admin, who has been around since at least 2003. This admin is currently active, so this is not them, but it's an interesting coincidence. — Maile (talk) 19:47, 22 May 2019 (UTC)

  @Maile66: Not a coincidence - the logs were unhidden. Nv8200pa created the account and added the rights --DannyS712 (talk) 19:48, 22 May 2019 (UTC)

• Since it's been unhidden, I'll do it here: @Nv8200pa:. Can you explain this mess and also your unblock of this with the explanation of: (Incorrect ISP Identification; No longer owned Webhost; Residential IP Pool.), which is easily identified as a webhost. Followed by your assignment of IP block exempt to yourself via a UTRS ticket: Nv8200pa talk contribs changed group membership for Nv8200pa from administrator to IP block exempt (temporary, until 02:00, 22 May 2020) and administrator (Requested via UTRS #223560 - Trusted user across multiple projects)) (thank)? And can another administrator with UTRS access verify that it is indeed a ticket from the person in question? Praxidicae (talk) 19:49, 22 May 2019 (UTC)
  • 223560 isn't even a valid UTRS#, UTRS is only in the tens of thousands right now, not the 200 thousands. — xaosflux ^Talk 20:09, 22 May 2019 (UTC)

I'd also suggest a block of both accounts until this is sorted out as it appears to either be compromised or an egregious abuse of tools (or possibly incompetency) and none of these things are good. Praxidicae (talk) 19:51, 22 May 2019 (UTC)

• Praxidicae, See my comment here. The UTRS ticket has absolutely nothing to do with the admin, or IPBE. SQL^{Query me!} 19:51, 22 May 2019 (UTC)

  Looks like now blocked per CU as a compromised account. SQL^{Query me!} 20:00, 22 May 2019 (UTC)

  @SQL: Should this discussion be moved somewhere else? VPT probably isn't the right place --DannyS712 (talk) 20:03, 22 May 2019 (UTC)

  DannyS712, Probably could use a note at ANI. SQL^{Query me!} 20:11, 22 May 2019 (UTC)

  DannyS712, Done, see Wikipedia:Administrators'_noticeboard/Incidents#Probable_compromised_admin_account. SQL^{Query me!} 20:13, 22 May 2019 (UTC)

• Anyone thinks this is way too much technical proficiency for a compromised account? Deleting the main page for the lulz is one thing but this one made another account and redacted logs, used UTRS, added rights to themselves and an alt account. --qedk (t 桜 c) 20:04, 22 May 2019 (UTC)

Qedk Not at all. In one of the last spates of compromised admin accounts, they created alt accounts after editing innocuously and gave them perms too. If this was indeed a compromise, the EFM alone could do a TON of damage. Praxidicae (talk) 20:06, 22 May 2019 (UTC)

@QEDK: ^^^ --DannyS712 (talk) 20:07, 22 May 2019 (UTC)

This one tops that too is what I'm trying to say. Giving alts perms is one thing but erasing your tracks so meticulously is new to me. --qedk (t 桜 c) 20:09, 22 May 2019 (UTC)

QEDK, wasn't meticulous enough, DannyS712 caught it! SQL^{Query me!} 20:10, 22 May 2019 (UTC)

• @SQL: Yeah, before the logs were unhidden here is what I pieced together (I sent it to Praxidicae becauce I thought they were an admin and could check, but here it is verbatim):

Here is what I've managed to piece together:



From the user table in the database,

"user_id"       "       user_name"      "       user_registration"

36574779        "       NVPA8200"       "       20190522020807"



NVPA9800 was created on 05/22/2019 at 02:08:07



Log event (revision deleted) at 02:08, 22 May 2019 (likely creation of the account):

https://en.wikipedia.org/w/index.php?title=Special:Log&logid=99346409



Log event (revision deleted) at 02:09, 22 May 2019 by Nv8200pa (likely adding the user rights):

https://en.wikipedia.org/w/index.php?title=Special:Log&logid=99346427



New users (account creation) log item hidden at 02:11, 22 May 2019 by Nv8200pa:

https://en.wikipedia.org/w/index.php?title=Special:Log&logid=99346469



User rights log item hidden at  02:10, 22 May 2019 by Nv8200pa:

https://en.wikipedia.org/w/index.php?title=Special:Log&logid=99346460



I'm all for assuming good faith, and I could be completely off base since I can't see the log entries myself, but to me it appears that Nv8200pa created NVPA8200, gave them rights, and then hid it. You may want to check those log entries.

If you know what you are looking for, its not hard to piece together --DannyS712 (talk) 20:14, 22 May 2019 (UTC)

@DannyS712: Yeah but how'd you find NVPA8200? --qedk (t 桜 c) 20:19, 22 May 2019 (UTC)

@QEDK: I was bored DannyS712 (talk) 20:21, 22 May 2019 (UTC)

--qedk (t 桜 c) 20:24, 22 May 2019 (UTC)

@QEDK: In all seriousness, in the context of exploring @SQL: (the language) I ran a database query that I've recreated at Quarry:query/36356, ordering template editors by time of account creation. I ran similar queries with a few other rights, and thought I would investigate why an account created earlier today had all of these rights. I posted here, and then went on to figure the rest out, before Tony undeleted the lock entries so I could confirm my suspicions (the text above was sent when it was still hidden) --DannyS712 (talk) 20:28, 22 May 2019 (UTC)

Ahhh, so we got lucky afterall. But it's a good thing that we did. Let me make some progress regarding this over at ANI. --qedk (t 桜 c) 20:32, 22 May 2019 (UTC)

• Hi all, I ran a check after unrevdeling the log entries. There were several oddities in the edits that to me suggested the likelihood of account compromise. The account moved from editing on a static IP with one operating system for all previous edits within the CU period to a proxy with an entirely different device for the log entries in question. I have blocked locally, notified ArbCom, and requested a steward lock the account. TonyBallioni (talk) 20:12, 22 May 2019 (UTC)

  That was quick, @TonyBallioni:! --qedk (t 桜 c) 20:13, 22 May 2019 (UTC)

Hi. This morning in the UK I noticed that I was unable to access en.wikiscan, and that's still the situation this evening. I can access the site http://wikiscan.org/, but not http://en.wikiscan.org/ For the latter, I either get "The connection has timed out" error in Firefox or simply "Sorry, the website en.wikiscan.org cannot be found" via by ISP. Everything seemed OK yesterday. Thanks. Lugnuts ^{Fire Walk with Me} 18:01, 23 May 2019 (UTC)

In the US the same thing is occurring as well. OkayKenG (talk) 21:33, 23 May 2019 (UTC)

Appears to be working now. Thanks to whoever helped behind the scenes. Lugnuts ^{Fire Walk with Me} 13:34, 24 May 2019 (UTC)

Idk if this has been asked before, i first questioned this on the Teahouse and they redirected me to this page, i went to a band's page and it shows that they have a timeline but there is nothing on it so i tried editing to fix it, on the preview it is shown but on the real thing nothing (I use a phone to edit btw, so it's probably because of that). --FromFrank✎Talk♬ 14:35, 24 May 2019 (UTC)

See phab:T216318. You have to use desktop view to actually see the timeline, at least for now. – Ammarpad (talk) 14:46, 24 May 2019 (UTC)

Has anyone ever seen an editor who's able to upload while blocked? If so, please visit Commons:COM:VP/T and offer an opinion. Commons:User:Drbaseball95 was indefinitely blocked in 2012 and hasn't been unblocked, but as you can see from Commons:Special:Contributions/Drbaseball95 (temporarily, until I delete them) or Commons:Special:Log/Drbaseball95, the account uploaded dozens of images in 2019. Nyttend (talk) 00:52, 25 May 2019 (UTC)

@Nyttend: The files were uploaded to this wiki and then imported to commons by User:Magog the Ogre. * Pppery * _survives 01:13, 25 May 2019 (UTC)

Thank you for helping me understand. I was under the impression that a little marker (like <at en:wp>) was added to an uploader's username when an import occurred in the last few months, so I rejected the idea of an import when it came to mind. Nyttend (talk) 01:26, 25 May 2019 (UTC)

I just visited my bookmarked site to check contributions by new users, but it is way too slow to respond for some reason; I reloaded the page a couple of times, even with a different ISP, but it didn't help.

The page is Special:Contributions; it responds instantly, but if I add these two parameters: contribs=newbie&target=newbies, it takes roughly 40 seconds to load, which is absurd.

If you'd like to help, please let me know if you can reproduce the issue: link with the said parameters added. Thanks. —RainFall 11:30, 17 May 2019 (UTC)

• @RainFall:  Confirmed I got this error: PHP fatal error: entire web request took longer than 60 seconds and timed out. Must be quite an intensive request. Guywan (talk) 12:12, 17 May 2019 (UTC)
• There's phab:T137478; filtering recent changes by new users is much faster ([17]). Galobtter (pingó mió) 17:25, 17 May 2019 (UTC)

The Special:Contributions&contribs=newbie&target=newbies worked fine until about 10 days ago when I started getting the same problem as User:RainFall. The phab task to remove this option as I see it is not a good idea despite the ability to get pretty much the same information via Special:RecentChanges. Nthep (talk) 17:51, 17 May 2019 (UTC)

Nthep, ok. so you have a patch to make the thing magically faster ? —TheDJ (talk • contribs) 21:27, 17 May 2019 (UTC)

I wish. Although the same link on Commons continues to be rapid. I accept that the load here is greater but something must have been changed to make the performance degrade so badly. Nthep (talk) 21:34, 17 May 2019 (UTC)

New user contributions slow

For the past week or so, whenever I try to look at new user contributions ("Show contributions for new users only"), the query takes VERY long (sometimes so long that the Wikimedia server gives up and throws up an error page). Has anyone else experienced this? Anyone looking into it? WikiDan61^ChatMe!_ReadMe!! 00:31, 24 May 2019 (UTC)

See #A "special" page is taking a very long time to load. PrimeHunter (talk) 08:47, 24 May 2019 (UTC)

Newbie edits function not working

Hi all,

In recent weeks, the new editors' contribs feature has not been loading. When it (rarely) does, it is very, very slow. I've also been getting Wikimedia errors ("Our servers are currently under maintenance or experiencing a technical problem.") Would appreciate any help resolving this.

Thanks,

GAB^gab 08:26, 25 May 2019 (UTC)

See #New user contributions slow and #A "special" page is taking a very long time to load sections above. – Ammarpad (talk) 09:13, 25 May 2019 (UTC)

Moved from Wikipedia talk:Village pump (technical) § Lead section paragraphs out of order again

 – --Redrose64 🌹 (talk) 13:57, 25 May 2019 (UTC)

At Surtsey on mobile, the second paragraph has been put above the infobox instead of the first one, resulting in an incorrect reading order. This isn't the first time this has happened. Hairy Dude (talk) 02:52, 19 May 2019 (UTC)

Hi Hairy Dude, next time you should post at the actual village pump Wikipedia:Village pump (technical) instead of this talkpage as people rarely reply to posts here. – Ammarpad (talk) 12:32, 25 May 2019 (UTC)

Wikimedia received an email from someone attempting to create a new username. ticket:2019052510000337 They were prompted to enter a Captcha with the phrase "nazisgood". They were understandably surprised as were some other agents, who thought this would be excluded by a blacklist.

I thought that there may be editors here who might be able to said some light on the issue.--S Philbrick(Talk) 14:34, 25 May 2019 (UTC)

I assume it is because "nazis" isn't in the bad word list (which does appear to exist per phab:T163148). Galobtter (pingó mió) 16:33, 25 May 2019 (UTC)

Galobtter, That sounds plausible. S Philbrick(Talk) 18:17, 25 May 2019 (UTC)

Was there a picture of the catpcha attached? Reedy (talk) 18:37, 25 May 2019 (UTC)

Reedy, there is now. S Philbrick(Talk) 21:38, 25 May 2019 (UTC)

Is there a way to tag an image in an article so that it won't be used as a thumbnail in mobile search? GoldenRing (talk) 16:03, 23 May 2019 (UTC)

It's called the page image and is used for several things. mw:Extension:PageImages#Can I exclude certain page images? says: "There is a blacklist on every wiki that administrators can edit, at MediaWiki:Pageimages-blacklist (example). Any images that appear here will not be page images for any article." I guess we shouldn't make a long MediaWiki:Pageimages-blacklist for efficiency reasons. Moving the image out of the lead would also work. PrimeHunter (talk) 16:49, 23 May 2019 (UTC)

Let's beat around the bush a little less and get to specifics - when searching on mobile with terms starting with "human" (hell, I even managed to reproduce it with "hum"), the pictures on human penis and human penis size show up as thumbnails. There is some concern that this is inappropriate if you're searching for something innocuous like the article on Huma Abedin or hummus. Primefac (talk) 14:55, 26 May 2019 (UTC) (please ping on reply)

At this Redirects for Discussion entry, concern was expressed over them cluttering the search results, as a justification for deletion. This seems like a kludge that does not address the underlying problem.

I'm writing here to open a discussion about the desirability/feasibility of marking redirects with, for example, Category:Unsearchworthy redirects, and having them be removed from the search dropdown, and from search results unless they're specifically requested to be included. That would allow the redirects to be retained to preserve inbound links if desired, without getting in the way. (It would make them WP:CHEAPer, I suppose.)

Thank you for considering this! —((u|Goldenshimmer)) (they/their)｜😹｜✝️｜John 15:12｜☮️｜🍂｜T/C 18:16, 23 May 2019 (UTC)

Isn't this already a feature of some templates, like ((R unprintworthy))? Perhaps Paine Ellsworth (talk · contribs) knows. --Redrose64 🌹 (talk) 19:15, 23 May 2019 (UTC)

Oh, that would be cool if so! —((u|Goldenshimmer)) (they/their)｜😹｜✝️｜John 15:12｜☮️｜🍂｜T/C 22:55, 23 May 2019 (UTC)

If so that is news to me (and that would be great)! ((R from subpage)) and ((R from camelcase)) both should have this feature, if it exists. Some of the RfD guidelines would also probably need to be updated if this feature exists. - Paul^T+/_C 03:09, 24 May 2019 (UTC)

I agree that this sounds very much like editors with an R#DELETE #1 agenda testing the meaning of "unreasonable" as it applies to search-field-dropdown-menu hogs, and I also agree that R#KEEP #4 far outweighs. I know of no redirect category for such hogs. And I wonder if anyone in that RfD has actually tested "The Simpsons" in their search field? It yields very different results than "The Simpsons/" with the backslash. Printability is an altogether different issue, yet there is likely some overlap. For the millions of mainspace redirects, one thing that unprintworthy redirects would have in common with unsearchworthy redirects is that this is not a choice that should be given to a bot, since it would often be difficult to assess the "unworthiness" of a redirect. This will remain an issue that should be decided at RfD. Paine Ellsworth, ed.  ^put'r there  08:49, 25 May 2019 (UTC)

That all makes sense. There are definitely some tricky edge cases with this feature that require human judgement. Do you agree that this is a feature that makes sense? Where is the best place to request it? WP:PHAB? Or does there need to be more discussion first, perhaps at WT:SEARCH? - Paul^T+/_C 12:24, 25 May 2019 (UTC)

To editor Paul: it certainly makes sense to minimize the search field noise if that's possible. As I wrote in my RfD rationale, when I type "The Simpsons" in my search field, I get very different results than when I type it in with the backslash, as in "The Simpsons/", so I don't see the problem there. It seems that all the results for "The Simpsons" have "natural" priority over the results for the backslashed version. I might not be the right person to ask. Couldn't hurt to write a phab ticket, because at least that way you get to find out from the devs whether or not it's a good idea or even doable. Want you to know that I appreciate your interest in all this, because it seems to me that filling and monitoring such a category would be a major task with a long-term commitment. Paine Ellsworth, ed.  ^put'r there  11:31, 26 May 2019 (UTC)

@Paine Ellsworth: This →/← is a slash. A backslash is this →\←. --Redrose64 🌹 (talk) 20:18, 26 May 2019 (UTC)

To editor Redrose64: lol, thank you! always getting those wards backwards. Paine Ellsworth, ed.  ^put'r there  20:29, 26 May 2019 (UTC)

On June 3 the _user and _user_text fields of revision, logging and a few other tables on the Toolforge replicas will be dropped, as part of the actor migration that has been going on for over a year. The full list of changes can be found at phab:T223406. At simplest, you'll need to JOIN on actor to get user names and/or the value for the *_actor foreign key in revision, etc. Pinging maintainers of bots and tools that I'm fairly sure are affected: @Enterprisey, Σ, SQL, and MZMcBride. The time to update your tools is now. If you're not already, you should subscribe to the Cloud-announce mailing list to get notified of such changes with more advanced notice. Best, — MusikAnimal ^talk 00:10, 21 May 2019 (UTC)

MusikAnimal, I've updated SQLBot's AIV task. I'll try to hunt it down in all my remaining tools when time permits. SQL^{Query me!} 06:13, 21 May 2019 (UTC)

Oh, yikes. Thank you for the ping! This change is almost certainly going to break some scripts that I'm foolishly still maintaining. --MZMcBride (talk) 07:11, 25 May 2019 (UTC)

Thanks for the ping! My stuff should be good but I'll do another check. Enterprisey (talk!) 21:38, 26 May 2019 (UTC)

I've had User:Ucucha/duplinks in my common.js for a very long time. It used to highlight duplicate links pretty well. Now, the only way I can tell what it's highlighted, if anything, is by enlarging my screen and using a magnifying glass. And how it highlights is one link is a a teeny-weeny very thin red line around one duplication, and a very faded dotted box around the other instance. For one thing, red is not easy to see for some visions that have no problem with other colors. And what is the use of making the other link so subtle that it can hardly be seen? Is there a better tool than this to highlight the duplicate links? This is just not workable as it. I've tried it on Firefox and Chrome, so the issue is not browser specific. — Maile (talk) 01:05, 25 May 2019 (UTC)

@Maile66: User:Ucucha/duplinks.js was merged with User:Evad37/duplinks-alt.js on 24 January 2019. Have you asked Evad37 (talk · contribs) about it? --Redrose64 🌹 (talk) 12:47, 25 May 2019 (UTC)

Redrose64 (talk · contribs) Thanks. I switched the script with Evad37's script. Same thing. Perhaps that's the editor who changed the visuals. I'll contact him. — Maile (talk) 12:57, 25 May 2019 (UTC)

Maile66 - it looks like the script adds two CSS selectors: .duplicate-link { border: 1px solid red; } and .duplicated-link { border: 1px dashed green; }. You could probably add those to your common.css and modify them to get thicker lines, other colors, or anything you might want. For instance, try this to make the lines 3 times as thick:

.duplicate-link {
    border: 3px solid red !important;
}
.duplicated-link {
    border: 3px dashed green !important;
}

(note: I haven't tested this, and haven't used the extension, so I don't know if this would work) --rchard2scout (talk) 10:37, 27 May 2019 (UTC)

Color me confused.

A new editor user:Jflemmingjr wrote to Wikimedia, ticket:2019052510003218 saying they could not login. Their account is two days old with zero contributions. I didn't see any sign they were blocked, so I asked them to try again. It didn't work and they provided a screenshot.

Then I remembered I had run into a similar situation, and I was told to check to see if it was a global block.

I checked here, and see no sign of a block.

According to Wikipedia:Blocking_policy#Global_blocks:

Registered users cannot be globally blocked. The analogous action is global locking, which prevents anyone from logging into the account.

However, the screenshot clearly states that they are "blocked globally".

What am I missing?--S Philbrick(Talk) 16:13, 25 May 2019 (UTC)

The account Jflemmingjr is not globally locked and the only global blocks that exist are IP global blocks. Ruslik_Zero 20:39, 25 May 2019 (UTC)

Ruslik0, Thanks. I thought a message notifying an individual that they are subject to a global lock would identify the IP address. I saw a sample somewhere recently but can't seem to track it down at the moment, but the message this person received identify their username not their IP address so I'm still unclear on what happened. I've asked them to share their IP address so I can check the global lock list. I won't be terribly surprised if they failed to tell me their IP address. I know I wouldn't share mine if someone simply asked. S Philbrick(Talk) 21:51, 25 May 2019 (UTC)

• I was blocked without being blocked thanks to an IP block. See the threads on my talk page about it. It was highly distressing. DuncanHill (talk) 21:53, 25 May 2019 (UTC)

Pinging @Swarm: who helped me. DuncanHill (talk) 21:57, 25 May 2019 (UTC)

DuncanHill, the individual shared their IP address with me, which I'm not going to share publicly.

They created their username two days ago so I assume they want blocked or locked then.

I looked at all global blocks since then and did not see it in the list. I also looked to the last 500 to see if it was an arranged and it wasn't. The page has a box to enter the IP address which I did and the response was "The requested IP address is not blocked." S Philbrick(Talk) 22:38, 25 May 2019 (UTC)

One of the things which baffled everyone about my case was the message saying "you are blocked globally" when I wasn't blocked anywhere, that ties in with the initial post here. I don't recall the specifics of how I found out it was a global IP block, but it was bloody obscure. DuncanHill (talk) 22:54, 25 May 2019 (UTC)

Yes, it was a global IP rangeblock that was preventing a password change. We have a rangeblock finder around here somewhere... ~Swarm~ ^{sting} 02:56, 26 May 2019 (UTC)

The IP address associated with Jflemmingjr is indeed not globally blocked. So, it is not clear what it is all about. Ruslik_Zero 13:25, 26 May 2019 (UTC)

It would be helpful is the user showed a screenshot of the actual block message. Ruslik_Zero 13:29, 26 May 2019 (UTC)

They did send the screenshot in the ticket. The source message is MediaWiki:Globalblocking-blocked-nopassreset. – Ammarpad (talk) 13:38, 26 May 2019 (UTC)

It's possible that they're trying to log in through a blocked proxy or VPN, but supplied their actual IP address in the ticket. —DoRD (talk)​ 13:48, 26 May 2019 (UTC)

• @Sphilbrick: - yes, the message Ammarpad links to is precisely the situation that DuncanHill encountered. Due to a bug, underlying global IP blocks prevent password changes, even if set to anon-only, and even for accounts with IPBE (there are multiple Phab tickets about this languishing in the backlog). If you can't find the block, it is likely because it is a rangeblock. Rangeblocks can be tracked down using Rangeblock finder and disabled locally, even if just temporarily, to allow the blocked user to regain access to their account. That is currently the only workaround that I am aware of. It's good that DuncanHill saw this because this is an obscure one. ~Swarm~ ^{sting} 18:39, 26 May 2019 (UTC)

  Swarm, Thanks, although apologies in advance for not being sure how to carry out the next step. The individual informed me that they almost always use their phone. They did give give me an IP address but it's an IPv4. For some reason, I thought that mobile devices was used IPv6. Is that assumption flawed?

  I did enter the IP into the range finder. Both the local block count and the global block count are identically zero for all entries. My interpretation is that there is no block for this IP address. Is that correct? S Philbrick(Talk) 19:14, 26 May 2019 (UTC)

  My mother's iPad connects to Wikipedia via the WiFi router in her house. Seen from outside (i.e. from Wikipedia), her iPad's IP address is actually that of the router; being dynamic, it's pot luck, and in my experience is often IPv4. --Redrose64 🌹 (talk) 19:45, 26 May 2019 (UTC)

  No, that assumption is indeed flawed. As Redrose says, if I use my phone at home it will reflect the router's IP (IPv4). If I go to the pub across the road it will reflect the pub's router (also IPv4). However if I turn off WiFi it will pick up the local 4G signal for my phone's provider ... which is also IPv4. This is in the UK, it may of course be different for other countries. Black Kite (talk) 21:44, 26 May 2019 (UTC)

  IPv4/IPv6 is about allocation, since IPv4 addresses are considerably old, more fixed-line ISPs opt for IPv4 addresses since IPv6 is relatively new and lots of old fixed-line devices don't support IPv6. Phone service providers take more IPv6 since there's more of them available and newer phones support any technology you throw at them. --qedk (t 桜 c) 07:22, 27 May 2019 (UTC)

  Black Kite, thanks for cleaning up my flawed assumption S Philbrick(Talk) 14:56, 27 May 2019 (UTC)

15:33, 27 May 2019 (UTC)

Why do Special:Diff and Special:MobileDiff react differently to an invalid ID number when I load them on a laptop? They work similarly with a valid ID; https://en.wikipedia.org/wiki/Special:Diff/898964501 and https://en.m.wikipedia.org/wiki/Special:MobileDiff/898964501 both take you to an edit I made a few minutes ago, and there aren't any differences beyond what you expect with mobile versus desktop. However, https://en.wikipedia.org/wiki/Special:Diff/898964501898964501 and https://en.wikipedia.org/wiki/Special:MobileDiff/898964501898964501 do not react the same; Diff gives you MediaWiki:Difference-missing-revision, just as it would if you linked to a diff that's since been deleted, but MobileDiff produces a 404 error. I can't envision why it doesn't return some sort of MediaWiki error page (or take you to the Main Page) rather than a 404 error; this is the first time I can ever remember getting a 404 error on any WMF page when the servers were working properly. Nyttend (talk) 01:31, 27 May 2019 (UTC)

I observed this issue for a while but didn't think it's enough problem to report. It's still low priority, nice to fix. Reported as phab:T224430. – Ammarpad (talk) 15:37, 27 May 2019 (UTC)

Seems most noted wikilinked items in the ((Life timeline)) template work very well - but (using Dell Wintel10-1903 PCs/Chrome & Firefox browsers) several links seem to have stopped working well for some reason, including the "Flowers", "Plants", "Arthropods", "Birds" and "water" links - fixes or workarounds welcome - in any case - Enjoy! :) Drbogdan (talk) 13:48, 27 May 2019 (UTC)

This template is used in 136 pages. Please revert to the last working version and experiment in the sandbox. Thanks. – Jonesey95 (talk) 17:43, 27 May 2019 (UTC)

@Jonesey95: Thank you for the advice - restored the last working version to the template - all links now seem to be working ok - seems some code (re: "lint errors resolved"?) may have been added which seems to have been the problem - Thanks again - and - Enjoy! :) Drbogdan (talk) 18:53, 27 May 2019 (UTC)

According to MOS:DOUBLE SPACE, "Some editors type two spaces after a period/full stop; these are condensed to one when the page is rendered, so what the reader sees is not affected". This has always been my experience. However, today I saw that double spaces in wikitext are showing up as double spaces in rendered articles. I first noticed the issue here; see the double spaces in the first paragraph. The problem is present in Firefox 67, Chrome 74, and Internet Explorer 9 on my Windows 7 machine. Has a MediWiki software change caused this problem? Can it be fixed? Thanks! --Albany NY (talk) 03:19, 28 May 2019 (UTC)

That article had a lot of non-breaking spaces (hex 00A0). That is why some double spacing was visible. It's been cleaned up now. Johnuniq (talk) 03:30, 28 May 2019 (UTC)

Oh good, I'm glad it's not a widespread problem. Thanks for your quick response! --Albany NY (talk) 17:12, 28 May 2019 (UTC)

If you use Timeless, you may have noticed the background image, the angry cat chewing its way out. As I announced in the newsletter, this can now be configured to be, well, whatever, by setting it to a different image. So two questions:

1. Does anyone care what we put there, or should I just upload a drunk patch with a kitten rolling around with the puzzle globe like a ball of yarn at 2am sometime?
2. Where should the image live? Is it better to have the file(s) locally or on commons and protected such that only admins can overwrite them, or have them with the rest of the wmf-config images, where developers need to submit patches for review to overwrite them?

That last bit may also apply to the logo itself. There is a chance we may be wanting to start migrating the logo in all skins into a two-part thing, with the wordmark separate from the logo image - maybe sometime in the next 5-10 years - and this is something I intend to test in Timeless in particular soon. Given that it would be better to be able to reuse the Minerva wordmark for that one, though, we will probably want all of that in the config repo...

Any thoughts? -— Isarra ༆ 09:38, 27 May 2019 (UTC)

@Isarra: I rarely use timeless, but look at it occasionally. I'm creating this reply in timeless right now to check, I'm not getting any background image at all. Are there pages we should be seeing it on? — xaosflux ^Talk 15:12, 28 May 2019 (UTC)

@Xaosflux: It's there, just not obvious (like I didn't know there was a book in Monobook). Look at short pages such as homo (subgenus) and see the curves just below the bottom of the content-proper. --Izno (talk) 15:21, 28 May 2019 (UTC)

@Izno: thanks, I'm able to see just the very bottom edge of it sticking out. So as far as this goes, I suggest: Change to blank / disable, since this is a reader-facing easter-egg that is inconsistently displayed (based on body size) and doesn't contribute to the articles. — xaosflux ^Talk 15:25, 28 May 2019 (UTC)

You can see it more clearly on Special:BlankPage. I personally like it, and I think a similar watermark-like variant of the WP logo would make more sense. But, it is sort of an easter egg that doesn't hold much value. — MusikAnimal ^talk 17:18, 28 May 2019 (UTC)

I have an occasional problem I have been unable to reproduce. I normally work as a registered user in the mobile view (iOS, iPhone, Safari). Sometimes when I invoke the desktop view, I return to the mobile view by the back arrow (<) in Safari rather than the "Mobile view" link in the page footer. There have been times when I suddenly find myself logged out. I have also edited pages in desktop view and discovered that the edit was recorded with my IP address and that I had indeed been logged out at some point. I can also mention that when I log out in the mobile view and have invoked the desktop view in a session, I am switched to the desktop view with a session error message. I then retry and am logged out normally. This behavior is consistent. I have seen the FAQ entry "Hey! Why was I automatically logged out?", which might apply. Jmar67 (talk) 11:21, 29 May 2019 (UTC)

At https://en.wikipedia.org/w/index.php?title=Ellen_DeGeneres&oldid=899325737#External_links, the bulleted list seems like it's way off to the left several pixels and the bullets seem very large--I have used this site for over 10,000 hours for over 15 years... Am I just seeing things? Is someone monkeying with the site-wide CSS somehow? e.g. compare with https://en.wikipedia.org/w/index.php?title=Chronic_Town&oldid=898933556#External_links, where the bullets are steely-blue squares that are several pixels to the right rather than black dots to far to the left. ―Justin (koavf)❤T☮C☺M☯ 09:25, 29 May 2019 (UTC)

It looks like it might be to do with the presence of ((Wikinewspar2)) in that section - preview with that template removed and the bullets look like you'd expect to see. Nthep (talk) 09:36, 29 May 2019 (UTC)

Apparently ((Wikinewspar2)) does something weird with the CSS for bulleted lists. Adding an empty line between the template and the start of the list fixed it. rchard2scout (talk) 10:55, 29 May 2019 (UTC)

I've updated Module:Side box to make a better fix. -- WOSlinker (talk) 11:26, 29 May 2019 (UTC)

Is Earwig's tool broken? I can't get it to give me anything but 500 Internal Server Error. Justlettersandnumbers (talk) 12:02, 29 May 2019 (UTC)

Justlettersandnumbers, It works for me. I was getting errors on some of my tools earlier this morning, don't know if that might be related. SQL^{Query me!} 13:16, 29 May 2019 (UTC)

Yes, it seems to be back now. Thanks, SQL, Justlettersandnumbers (talk) 13:36, 29 May 2019 (UTC)

What language is this?

𑢹𑣈𑣚𑣚𑣉𑢿𑣉𑣜𑣚𑣑

All I see are boxes. I found this under Asian input methods. Guywan (talk) 20:02, 29 May 2019 (UTC)

It would seem to be from Warang Citi (Unicode block), which is used for the Ho language (insert bad joke here). –Deacon Vorbis (carbon • videos) 20:10, 29 May 2019 (UTC)

@Guywan: (forgot to ping). –Deacon Vorbis (carbon • videos) 20:17, 29 May 2019 (UTC)

The script is Warang Citi [20], one of the writing systems for Ho language. Maybe your browser shows boxes with the hexadecimal values of the characters (my Firefox does that). What do you mean by "under Asian input methods"? If it's a page then always link pages you refer to. PrimeHunter (talk) 20:21, 29 May 2019 (UTC)

@Deacon Vorbis: Ho'ly moly, thanks a ho lot.

Not a page. Input methods via the cog in the sidebar. Guywan (talk) 20:53, 29 May 2019 (UTC)

@PrimeHunter: Forgotten ping. Guywan (talk) 20:57, 29 May 2019 (UTC)

Firefox does show boxes with the hexadecimal values of the characters; Opera just shows empty boxes.

Raw characters may be converted into their Unicode representations using Richard Ishida's Unicode code converter. Paste the mystery chars into the green area near the top, and click the "Convert" button above that. Various representations will appear in the grey areas. Just above the first grey area, titled "Characters", is a button "View in UniView", this is a direct link to the tool that PrimeHunter mentioned.

Another way of reaching this "Input methods" thing is via Special:Preferences: just below the gender selection and above the custom signature settings is a link "More language settings". Unfortunately, what gets displayed in this popup is dynamic and depends upon a number of factors. English is always written using the Latin script, so you won't get input methods for English variants. --Redrose64 🌹 (talk) 23:02, 29 May 2019 (UTC)

Chrome also happens to display these as boxes. --Izno (talk) 23:44, 29 May 2019 (UTC)

Well, strictly speaking, browsers only show boxes if you don't have an appropriate font installed. I see the characters, though they are rather blocky—I think they are being rendered with GNU Unifont (which substitutes completeness for beauty). Eman235/talk 00:32, 30 May 2019 (UTC)

Resolved

I'm not seeing this on en:wp yet but I am at Commons and Wikidata. On my contributions page this list of accounts show automatically. I can vaguely see a usefulness in allowing quick selection of alternate accounts (Nthep-vanilla is an alt account of mine) but seeing another username which is nothing to do with me is annoying. Why can't it be a user operated dropdown rather than being automatic and how should the accounts displayed be controlled? The layout is also pretty rubbish overlaying other fields like that. Is this a recent change to Mediawiki? Nthep (talk) 11:51, 30 May 2019 (UTC)

Having it auto expand is bad at best, someone needs to open a phab on this. The auto selector should only apply when clicking IN TO that field, not just because it is auto-selected. — xaosflux ^Talk 12:10, 30 May 2019 (UTC)

Happens in other dialogs like the global user rename as well. –xeno^talk 12:16, 30 May 2019 (UTC)

Happens with all Performer/Target text boxes in Special (Contributions included), I don't see a problem with it. With a decent internet connection, the lag is just barely noticeable. --qedk (t 桜 c) 12:30, 30 May 2019 (UTC)

This will soon be resolved at phab:T224524. – Ammarpad (talk) 12:33, 30 May 2019 (UTC)

QEDK, I'm not sure what you mean - the list doesn't go away unless I make a deliberate and previously unneeded selection. Neither does the list disappear if I make any other field the focussed field. Nthep (talk) 12:48, 30 May 2019 (UTC)

@Nthep: It seems that the bug was that it was automatically brought up expanded, I couldn't replicate the issue so I assumed you were talking about the autocomplete suggestions being annoying. Either way, it's fixed now per the Phab ticket (and my Contributions looks ok as well). --qedk (t 桜 c) 13:49, 30 May 2019 (UTC)

So sometime over last night my watchlist behavior changed, and the nodes on the side won't change from green to grey after I visit the diffs like they usually do. The marked pages as read seems to work, though. I do have some custom CSS which I know might be an issue, and am using Monobook, but it is extremely useful to know if I missed anything when I go through my list. ♫ Melodia Chaconne ♫ (talk) 12:38, 2 May 2019 (UTC)

Not caused by your custom CSS, being looked at in phab:T218511. – Ammarpad (talk) 14:35, 2 May 2019 (UTC)

I hope it's sorted soon. Today I saw a page on my watchlist with new edits not marked as new. It's the first time I've noticed this; it's generally been the problem described above. This behaviour's described on the phabricator page as well. BlackcurrantTea (talk) 04:05, 3 May 2019 (UTC)

I've been having this and related problems on and off since mid-March. See here for past discussion. I hope this all gets done with soon. —Granger (talk · contribs) 11:46, 3 May 2019 (UTC)

Yes, I'm seeing it again too now, with the added benefit of changed, unseen pages showing up without bold. It's kinda insane this is STILL an issue six weeks later! —Joeyconnick (talk) 03:42, 5 May 2019 (UTC)

• Sigh. And now bolded and non-bolded entries are entirely random. GMG^talk 12:18, 3 May 2019 (UTC)

Melodia, I'm getting pages showing I've visited since the last change when it's not true. --valereee (talk) 18:09, 4 May 2019 (UTC)

I get this too. – Tea2min (talk) 06:40, 5 May 2019 (UTC)

My watchlist is buggered too, exactly as described above, but that Phabricator report is weeks old, and my issue only started a couple of days ago.Roxy, the dog. wooF 14:27, 5 May 2019 (UTC)

It seems that it was related to whatever update included a time stamp on the "view new changes" button. Or maybe that's just a coincidence. At any rate, it's a problem across all projects, and not just the English Wikipedia. GMG^talk 14:46, 5 May 2019 (UTC)

The issue with Firefox certification made teh Internetz unusable for me yesterday. Mozilla script kiddies fixed that quite quickly. Could WMF poach some of the Mozilla nerds to help? Roxy, the dog. wooF 14:52, 5 May 2019 (UTC)

As noted above, the issue has been affecting some users (including me) since March, so something else must be involved with the problem. isaacl (talk) 17:26, 5 May 2019 (UTC)

Started happening for me a few days ago too - David Gerard (talk) 17:36, 10 May 2019 (UTC)

Seems to have stopped happening for me now - David Gerard (talk) 19:35, 13 May 2019 (UTC)

Just started again for me. —Joeyconnick (talk) 05:23, 14 May 2019 (UTC)

I have these two gadgets enabled: "^(D) Display green collapsible arrows and green bullets for changed pages in your Watchlist, History and Recent changes (unavailable with the improved Watchlist user interface)"; "Display pages on your watchlist that have changed since your last visit in bold (see customizing watchlists for more options)". Both have been misbehaving for several weeks now. But interestingly, there was an occasion a day or two back when the Wikimedia servers were slow, and not all of the JavaScript and CSS was being sent back to me. During that period, the Watchlist was behaving as it should: unread posts were boldfaced with a green bullet; read posts were normal weight with a blue-grey bullet. Once the servers were back to speed, the misbehaviour resumed. I conclude that for each of the gadgets, there are two different scripts that have different effects that conflict with one another. --Redrose64 🌹 (talk) 16:21, 14 May 2019 (UTC)

Perhaps we'll find out later this week. According to the Phabricator page this week's MediaWiki release fixes it. In the meantime I'm thinking of my watchlist as one of those irritating online shops where they keep suggesting the same items, whether or not you've seen them ('Other editors who read Village pump (technical) were interested in...'). BlackcurrantTea (talk) 16:59, 14 May 2019 (UTC)

Still not fixed. Someone's posted about it on Phabricator; the task is still marked as resolved, and I can't tell if anyone's opened a new one. BlackcurrantTea (talk) 18:06, 17 May 2019 (UTC)

Yes, it's still broken. Read items still marked as unread. I don't know what the "Closed, Resolved" status means: either it's fixed in the code but not deployed on the servers yet, or it's already deployed, but then it means the fix does not fix it, and therefore the bug should be reopened.—J. M. (talk) 21:06, 18 May 2019 (UTC)

How is this still happening? More than two months and this problem is still showing up. This morning most of the pages on my watchlist aren't getting marked as read, and those that are get marked as unread again a few minutes later. —Granger (talk · contribs) 23:35, 23 May 2019 (UTC)

@Aaron Schulz: It looks like it was you who marked this as resolved on Phabricator, so I'm pinging you to let you know that the problem is still happening. —Granger (talk · contribs) 23:38, 23 May 2019 (UTC)

Is there a way to reset the entire watchlist? Geordie (talk) 23:49, 23 May 2019 (UTC)

Yes, there's a "Mark all changes as seen" button, if that's what you mean. —Granger (talk · contribs) 00:16, 24 May 2019 (UTC)

Where is this button to be found? Geordie (talk) 00:32, 24 May 2019 (UTC)

Thought it was fixed but just saw one of my own changes to a watched page show up as unseen on my Watchlist. sigh —Joeyconnick (talk) 23:43, 30 May 2019 (UTC)

I know I've run into this before, but don't remember the fix. I just accepted Draft:Architecture of Saudi Arabia, and got the totally unhelpful error message:

Error while saving Architecture of Saudi Arabia: {"edit":{"code":"abusefilter-warning","message":{"key":"abusefilter-warning-predatory","params":["Predatory open access journals",891]},"abusefilter":{"id":891,"description":"Predatory open access journals","actions":["warn"]},"info":"Hit AbuseFilter: Predatory open access journals","warning":"
\"Your	Warning: An automated filter has identified this edit as introducing references to a predatory open access journal. If you are confident that you want to cite this source anyway, please click 'Publish changes' again. Note that citations to predatory journals are routinely removed.
","result":"Failure"))

This doesn't tell me the one thing that would be useful: the name of the journal that triggered the filter. How do I figure that out? -- RoySmith (talk) 14:15, 31 May 2019 (UTC)

@RoySmith: from filter 891 it looks like you triggered based off of doi=10.2495 being used. — xaosflux ^Talk 14:20, 31 May 2019 (UTC)

This is listed as "10.2495: WIT Press" in the filter notes. — xaosflux ^Talk 14:20, 31 May 2019 (UTC)

Would it be possible to update the error message to include a link to Special:AbuseFilter/891? That at least would get you pointed in the right direction. -- RoySmith (talk) 14:23, 31 May 2019 (UTC)

@RoySmith: I don't think that would be useful for most casual editors, but perhaps a Wikipedia: or Help: page exists that explains this concept for editors, and perhaps a list - we could link it from the error message (MediaWiki:Abusefilter-warning-predatory) for that filter. Does anyone know of a good landing page (or want to make one)? — xaosflux ^Talk 14:26, 31 May 2019 (UTC)

Well, I agree that the link to the abuse filter might not be the most useful thing, but it's certainly better than what we've got now, which is basically a message that says, "You did something wrong". There's 18 references in that article. Figuring out which one triggered the filter is essentially impossible from the information given. Actually, the best thing would be the text of the reference, plus a link to the filter. Then the user would know what was disallowed, and why it was disallowed. In other words, it makes the error message actionable. -- RoySmith (talk) 14:33, 31 May 2019 (UTC)

The edit filter can not inject variables in to the warning/action-block message, and I can't see us making dozens of separate filters (one for each predatory journal) just for that purpose (not to mention that an edit could hit multiple of them at once). If you had a list of "These journals/DOI's" you could at least see if your diff contained anything on the list though? — xaosflux ^Talk 15:24, 31 May 2019 (UTC)

Having a list isn't really useful. There's presumably 100's of journals on that list. This article had 18 references. Searching for the cross-hits between them is intractable. Sorry to be hard-nosed about this, but the current situation is pointless. You get an error message that doesn't give you enough information to fix the problem, and the operation completes anyway. -- RoySmith (talk) 15:57, 31 May 2019 (UTC)

phab:T216001 is the task regarding variables in warning messages; the list is actually only about ten or so long, so I've added the list to the warning message.

The filter is mainly for people who are adding content to existing articles or creating articles (not for draft to article moves etc by someone who didn't write the article), so it should be easy enough now for those people to figure out which journal triggered the filter. Galobtter (pingó mió) 16:52, 31 May 2019 (UTC)

Thanks for updating the message. That looks like it would be good enough to have resolved the problem on my own. My apologies if I sounded exasperated in the above thread. As a software developer, this is one of my pet peeves. Writing good error messages is not easy; they are often passed along from layer to layer and by the time they get to the end user, any useful context has been lost. So it becomes really important to make sure the message contains all the information the end user might need to resolve the problem. Ah, for the good old days when "?" was the only error message you needed :-) -- RoySmith (talk) 17:37, 31 May 2019 (UTC)

I've been trying to access "The Holocaust --> Page information --> Edits by user" for the last couple of days. The error message is long; it ends with "OSError: [Errno 116] Stale file handle". I've just tried a few other articles and got the same error message. Can anyone advise? SarahSV ^(talk) 00:25, 31 May 2019 (UTC)

The tool that that link calls is maintained by Σ, so pinging them. In the meantime, the same information is available via a different method as one of the External tools at the top of the history page. * Pppery * _{it has begun...} 00:35, 31 May 2019 (UTC)

Pppery, thank you. I've been able to access "edits by user" via the External tools. SarahSV ^(talk) 02:25, 31 May 2019 (UTC)

This is related to #Major schema changes coming to database replicas (permalink). There was an indexing change that went live a few days ago in preparation for this. That change didn't break older queries, rather made them really slow, I guess too slow in this case. I have updated MediaWiki:Pageinfo-footer to link to XTools TopEdits for now.

On Monday the schema change will be completed and Σ's User Search tool, among others, will no longer function. If a tool you use breaks, and the tool involves getting data about users in any way (as simple as displaying usernames), it's probably because of the schema change. You can direct maintainers to phab:T223406. — MusikAnimal ^talk 02:38, 31 May 2019 (UTC)

MusikAnimal, thanks for explaining and for changing the link. SarahSV ^(talk) 21:39, 31 May 2019 (UTC)

It used to be that when I searched Wikipedia, after I found an article, edited it and then returned to the search screen, the cursor would be positioned at that article. Now it goes back to the top of the screen, which makes me a curser. Is there any way to restore the original behavior? Clarityfiend (talk) 07:35, 31 May 2019 (UTC)

What is your browser? Internet Exploder has an annyoying habit of taking you right to the top when you return to a page that you had scrolled down in. --Redrose64 🌹 (talk) 13:15, 31 May 2019 (UTC)

I'm using Chrome on Windows 10. Clarityfiend (talk) 07:19, 1 June 2019 (UTC)

I always use the feature of my browser that opens a new tab, leaving the existing tab intact. In Firefox for Mac, for example, command-click any link to open the resulting page in a new tab, then command-option-right arrow takes you to that tab. You can use this feature to open a bunch of tabs at once, then work on them one by one, and eventually return to your starting tab. – Jonesey95 (talk) 15:57, 31 May 2019 (UTC)

You can disable Advanced Search in preferences (Search tab).

Maybe someone could add there also an option to disable autofocus? Or someone know how to disable/prevent autofocus in js script? MarMi wiki (talk) 23:14, 31 May 2019 (UTC)

Why autofocus is bad (after displaying the search results). MarMi wiki (talk) 23:35, 31 May 2019 (UTC)

Resolved

I'm trying to transclude a section set up with <section begin=One /> and <section end=One />, which is located at Arrow (season 1), but it doesn't work. I moved the same code to my sandbox and it did. Could there be an issue where this does not work for pages with disambiguation? --Gonnym (talk) 10:44, 30 May 2019 (UTC)

• @Gonnym: I meddled around and came up with the same result. I thought it was something to do with the table syntax but it works just fine in the user sandbox. But, when I was trying to use ((#lstx:)) (transclude article except section) I realised that the article was only transcluding the "Episodes" section, and therein lies the culprit, the usage of <onlyinclude> tags, that prevents the transclusion of the defined sections, as it's not allowed. You will need some work to remove the tags and bring the referencing articles into line. Sorry about the experimentation in and around your sandbox, I was trying to use one of the sandboxes as control. --qedk (t 桜 c) 11:18, 30 May 2019 (UTC)
  • Thanks for clearing that up! I'll see how I can fix that. And no worries about the sandbox, that's what it's meant for. --Gonnym (talk) 11:21, 30 May 2019 (UTC)
    • Just tested it out and it indeed works. --Gonnym (talk) 11:26, 30 May 2019 (UTC)
      • That's good to hear! --qedk (t 桜 c) 11:32, 30 May 2019 (UTC)
• And this is (part of) why I have been crusading against <onlyinclude> tags since July 2016. * Pppery * _{it has begun...} 11:36, 30 May 2019 (UTC)

  I'd rather campaign against most use of section transclusion, it is messy and unfriendly to new editors. — xaosflux ^Talk 12:12, 30 May 2019 (UTC)

  I agree that section transclusion is guaranteed to cause confusion and to periodically break when edits are made. Johnuniq (talk) 23:34, 30 May 2019 (UTC)

  Section transclusions would prevent most issues caused due to <includeonly> family of tags. Contrary to your supposition that section transclusions break things, it is the only transclusion tag that does not affect the entire page on which it is used (and allows multiple tags to nest or rest within each other with no issues) and thus, the best option for any kind of article transclusion. --qedk (t 桜 c) 07:25, 1 June 2019 (UTC)

Can someone please point me to a list of article redirects having categories. Also should they not be rid of categories? Capankajsmilyo^{(Talk | Infobox assistance)} 09:41, 1 June 2019 (UTC)

Category:Wikipedia redirects contains list of various subcategories of redirects. For your second question, you should ask at Wikipedia talk:Redirect or other appropriate venue as it's beyond the scope of this village pump.– Ammarpad (talk) 12:26, 1 June 2019 (UTC)

@Capankajsmilyo: There is no prohibition on categorising redirects - see Wikipedia:Redirect#Categorizing redirect pages. Previous discussion on this matter may be found at Wikipedia talk:Categorizing redirects/Archive 2#categorizing redirects as though they were articles and Wikipedia talk:WikiProject Shipwrecks#Categorization issue (look for Honey Lantree in both). --Redrose64 🌹 (talk) 21:52, 1 June 2019 (UTC)

It's annoying. Is there a fix?— Vchimpanzee • talk • contributions • 21:19, 17 May 2019 (UTC)

@Vchimpanzee: Go to the beta features page in Special:Preferences and verify that a) "enable all new beta features by default" is turned off and b) "use php7" is turned off. That aside, using PHP7 is not a bad thing, and there isn't really a reason for you to be annoyed. --Izno (talk) 21:25, 17 May 2019 (UTC)

Izno, the beta mode has just been removed, and now gradually more and more people will be switched over. —TheDJ (talk • contribs) 21:26, 17 May 2019 (UTC)

:O --Izno (talk) 21:27, 17 May 2019 (UTC)

Vchimpanzee, everyone is slowly being moved over to PHP7. This tag will be in place until everyone is moved over to php7. So this means you are in the group of the lucky few who have already moved over. Nothing to worry about. —TheDJ (talk • contribs) 21:25, 17 May 2019 (UTC)

As long as the tag stops appearing eventually.— Vchimpanzee • talk • contributions • 21:39, 17 May 2019 (UTC)

What is "php7"? DuncanHill (talk) 23:01, 17 May 2019 (UTC)

See PHP. -- WOSlinker (talk) 23:06, 17 May 2019 (UTC)

I did and it seems to be written for people who already know what the article is telling them, rather than for people who don't. DuncanHill (talk) 23:08, 17 May 2019 (UTC)

mw:Beta Features/PHP7#I am a non-technical user. What does this mean?: "Hopefully nothing. This is about changing the software used to run the wiki's code. Ideally, you should notice no real difference." PrimeHunter (talk) 23:44, 17 May 2019 (UTC)

MediaWiki, the software that runs Wikipedia, is written in a language called PHP. For a while, there were two ways for servers to run the PHP code: the default engine called Zend and an alternate method called HHVM. Wikimedia used HHVM because it was faster. A few years ago, HHVM decided that it would only support a specific style of PHP after PHP version 5. MediaWiki is not written in that style, so the developers decided to switch back to Zend. Zend had also become faster than HHVM. This change was first made as a beta feature to make sure nothing broke, but is now being rolled out on a larger scale to test it with more users. TL;DR: What PrimeHunter said. --AntiCompositeNumber (talk) 23:55, 17 May 2019 (UTC)

Thanks, that's the sort of explanation I was looking for. DuncanHill (talk) 23:58, 17 May 2019 (UTC)

@Vchimpanzee and DuncanHill: If you look at the history of many pages (such as this one), the contribs of some users like Vchimpanzee (talk · contribs), your warchlist or indeed this diff, you'll see that "(Tag: PHP7)" has two links. Click the second one. --Redrose64 🌹 (talk) 18:39, 18 May 2019 (UTC)

My contributions show that any edits I made at a library yesterday or Wednesday do not have the PHP7 tag.— Vchimpanzee • talk • contributions • 17:16, 31 May 2019 (UTC)

Tag: PHP7

Why are edit contributions being marked with Tag: PHP7? GoodDay (talk) 01:13, 24 May 2019 (UTC)

GoodDay, See this thread a bit up the page. --AntiCompositeNumber (talk) 01:17, 24 May 2019 (UTC)

PHP7 tag?

What is going on with the "PHP7" tag/filter on my recent edits, such as this one? Am I doing something odd/wrong? Steel1943 (talk) 22:45, 29 May 2019 (UTC)

@Steel1943: See Wikipedia:Village pump (technical)#All my contributions today say "Tag: PHP7" above DannyS712 (talk) 22:47, 29 May 2019 (UTC)

PHP7 tag doesn't show for edits made on some computers

No one responded when I posted here.— Vchimpanzee • talk • contributions • 20:34, 31 May 2019 (UTC)

Er, lots of people responded, myself included. Also, please don't start new threads for ongoing issues. --Redrose64 🌹 (talk) 21:31, 31 May 2019 (UTC)

No, not one response. This is a different question. There was no "Tag: PHP7" with any of my edits yesterday or Wednesday. It is related, but no one responded so I had to start a new thread.— Vchimpanzee • talk • contributions • 21:41, 31 May 2019 (UTC)

Did you follow the link that I provided at 18:39, 18 May 2019 (UTC)? Whether you are using PHP7 or not should make no difference at all. If there is some kind of difference, it's for the devs to worry about, not you or me. --Redrose64 🌹 (talk) 22:27, 31 May 2019 (UTC)

I don't see any explanation in what you linked to, but shouldn't all my edits be tagged with PHP7?— Vchimpanzee • talk • contributions • 16:08, 1 June 2019 (UTC)

It's being rolled out slowly. Currently ~10% of edits are randomly assigned to a php capable server. It's nothing to be concerned about, and you can safely ignore the tags :) Technical details for anyone interested: phab: T219127, mw:Topic:V0lqc8wseu459e0a -FASTILY 06:15, 2 June 2019 (UTC)

AFD stats: This is the path, and the message I'm getting: https://xtools.wmflabs.org/afdstats "File not found. The server said: No route found for "GET /afdstats"" The AFDstats issue has been happening for a few days. I posted over at Enterprisey, and it seemed to work for a different individual. I tried it on two different computers, two different browsers. Same results with all.— Maile (talk) 20:37, 1 June 2019 (UTC)

XTools is not the same as toolforge:afdstats. MusikAnimal maintains the former, and Enterprisey maintains the latter. Enterprisey probably needs to update the code for § Major schema changes coming to database replicas (T223406). — JJMC89 (T·C) 22:10, 1 June 2019 (UTC)

Enterprisey's tool works fine. This is MusikAnimal's domain (who I just pinged). You can consider raising this on Phabricator. --qedk (t 桜 c) 14:21, 2 June 2019 (UTC)

XTools has no AfD Stats feature, and as far as I know it never did. I'm not sure where Maile66 got that link. That said, toolforge:afdstats is still going slower than I recall it going before, which I assumed was because of the indexing change that happened last week. I made a pull request to fix it. If I'm right that those queries are still being used, the tool will break entirely tomorrow unless the fix is deployed. — MusikAnimal ^talk 15:57, 2 June 2019 (UTC)

MusikAnimal please see Enterprisey to do list. He developed it, and it's been around for years. He's the one you'd have to ask where it came from. — Maile (talk) 16:57, 2 June 2019 (UTC)

Ah, I knew putting it up on GitHub would be useful someday :) Thanks for the patch! I applied it, but the tool's still pretty slow. I'll see if there's anything else I can do. (As a credit-where-credit's-due sidenote, I've just been making minor fixes; Scottywong was the original author.) Enterprisey (talk!) 17:30, 2 June 2019 (UTC)

Hello.

Is there a suitable edit bot where you can submit edit suggestions for needed but tedious and repetitive work?

HandsomeFella (talk) 08:31, 3 June 2019 (UTC)

@HandsomeFella: This is really a question for WP:BOTREQ but they will ask for specifics, and may also direct you to read through WP:FDB and WP:CONTEXTBOT. --Redrose64 🌹 (talk) 08:34, 3 June 2019 (UTC)

• There's also Wikipedia:AutoWikiBrowser/Tasks if it's something which you think someone with WP:AWB might be able/willing to help you with. -- Begoon 08:45, 3 June 2019 (UTC)

Thank you both.

HandsomeFella (talk) 09:44, 3 June 2019 (UTC)

When editing, using both an alphabet list (e.g., Latin, Greek, or Cyrillic) and the "insert" list (long and short hyphens, etc.) at the bottom of the edit window, I am annoyed at having to switch back and forth between the two lists. Could the "insert" list (or at least its two hyphen variants) be included as an integral part of every alphabet list?

Thank you.

Nihil novi (talk) 13:38, 3 June 2019 (UTC)

How to disable (auto)focusing (page scroll to input) from search input in Advanced Search? This doesn't work (judging by the delay it seems that focusing is done somewhere in the AS code?):

// in common.js
var datao=JSON.parse($('#searchText').attr('data-ooui'))
delete datao.autofocus
$('#searchText').attr('data-ooui',JSON.stringify(datao))
$('input[type=search]').removeProp('autofocus')
$('#ooui-php-1').removeProp('autofocus')
$('input[type=search]').blur()
$('#searchText').blur()
$('#ooui-php-1').blur()

MarMi wiki (talk) 12:34, 1 June 2019 (UTC)

Okay, this seems to work:

$('#ooui-php-1').attr("disabled", true)

$.when(mw.loader.using("mediawiki.widgets.SearchInputWidget")).then( function(){
	$('#ooui-php-1').attr("disabled", false)
} );

Is there a better way? MarMi wiki (talk) 13:45, 1 June 2019 (UTC)

MarMi wiki, the latter isn't too good either. It messes with the loading order of the widget, by creating another loading point for it. Now whichever of the 2 comes first will be the one that is executed, which might lead to unintended consequences now or in the future. There really isn't a good way to do this right now honestly. There probably should be a searchAdvanced mw.hook at the end of the initialisation process for scripts to hook into. —TheDJ (talk • contribs) 08:59, 3 June 2019 (UTC)

It shouldn't be any difference as to when (true) module is loaded (unless code later modifies the loaded module) - that's why they have dependencies. And besides the module probably started loading itself before custom.js code is executing.

Revised code:

if (mw.config.get( 'wgPageName') == 'Special:Search')
{
  $.when(mw.loader.using("ext.advancedSearch.searchtoken")).then( function(){
  $('#ooui-php-1').attr('disabled',true)
  console.log('disable')
  //debugger
  //code from mw.inspect source
  //console.log(mw.loader.getModuleNames().filter( function ( module ) {return mw.loader.getState( module ) === 'ready';} ).join('\n'))
 } )

 //for testing purposes - what forces the focus
 $('#ooui-php-1').one("focus", function( event ) {
  //debugger
  console.log('0 '+$('#ooui-php-1').is(':focus'))
  $('#ooui-php-1').one("focus", function( event ) {
   console.log('1 '+$('#ooui-php-1').is(':focus'))
  })
 })
}

ooui-php-1 re-enables itself magically.

Choice of loader: from requirements to ext.advancedSearch.init (first forced focus in line $searchField.trigger('focus');).

Requirements checked in mw.inspect.getDependencyGraph()['module name'].

BTW, is it normal that methods of module aren't protected from invoking before needed module is loaded? Example: mw.inspect.getLoadedModules() fail with jQuery.Deferred exception: mw.inspect.getLoadedModules is not a function error, unless you invoke mw.inspect() first. MarMi wiki (talk) 15:19, 3 June 2019 (UTC)

15:24, 3 June 2019 (UTC)

If anyone familiar with quarry has a min can you take a look at quarry:query/18989? This used to run in about a min, now it is timing out for 30+ min run time. Thanks in advance :D — xaosflux ^Talk 15:38, 3 June 2019 (UTC)

Probably related to #Major schema changes coming to database replicas. --Edgars2007 (talk/contribs) 16:38, 3 June 2019 (UTC)

Hmmm, smells likely - if anyone feels like rewriting this for me go ahead :D — xaosflux ^Talk 19:57, 3 June 2019 (UTC)

@Xaosflux: Well, I have a few ideas.  Doing... --DannyS712 (talk) 19:59, 3 June 2019 (UTC)

@Xaosflux:  Done

• I removed the checks for transwiki-import (no users), import (just you and Graham, both of whom have other advanced rights that are checked), and epcoordinator (no longer in use)
• I joined the actor table per the change discussed above

See Quarry:query/36593 DannyS712 (talk) 20:11, 3 June 2019 (UTC)

@DannyS712: thanks, I'll have to go through my personal libraries and add that actor stuff in. — xaosflux ^Talk 20:14, 3 June 2019 (UTC)

Can somebody please help me implement an optional parameter for delinking ((Pre-nominal styles)) using ((Delink))? Thanks! Neveselbert (talk · contribs · email) 13:18, 3 June 2019 (UTC)

Not sure that I'd use ((delink)). Perhaps augment ((Pre-nominal styles)) with a |nolink= parameter which gets passed to ((Pre-nominal styles/GBR)) so perhaps this:

((for loop| |call=Pre-nominal styles/(({country|GBR))}|(({1|))}|(({2|))}|(({3|))}|(({4|))}|(({5|))}|(({6|))}|(({7|))}|(({8|))}|(({9|))}|(({10|))}|(({11|))}|(({12|))}|(({13|))}|(({14|))}|(({15|))}|(({16|))}|(({17|))}|(({18|))}|(({19|))}|(({20|))}|skipBlanks=yes|pc1n=nolink|pc1v=((#ifeq:(({nolink|))}|yes|1))))

If I understand how ((for loop)) works, the rather cryptic parameter |pc1n=nolink is the name of a static parameter that is sent to each call of ((Pre-nominal styles/GBR)) so in this case |nolink=; |pc1v= is the static value assigned to |nolink=, in this case when (({nolink|))} hold 'yes', |nolink= gets the value '1' else |nolink= gets no value.

Then in ((Pre-nominal styles/GBR))

| Canon = ((#if:(({nolink|))}|Canon|[[Canon (priest)|Canon]]))

When (({nolink|))} has a value, the switch returns Canon; when (({nolink|))} does not have a value, the switch returns [[Canon (priest)|Canon]]

Caveat lectur: not tested, but this is where I would start.

—Trappist the monk (talk) 14:05, 3 June 2019 (UTC)

Tweaked ((Pre-nominal styles/sandbox)) and ((Pre-nominal styles/GBR/sandbox)) (incomplete) and created minimal ((Pre-nominal styles/testcases)).

—Trappist the monk (talk) 14:46, 3 June 2019 (UTC)

 Thank you very much! Neveselbert (talk · contribs · email) 22:36, 3 June 2019 (UTC)

Is there a way to do this? If I search the Move log at Special:Log with Target set to an article name, it shows me when a page has been moved from that name, to a different one - which doesn't help me if I only know the current name, and want to follow the chain of renames backwards. Tried searching Hay's tools directory for terms like "move" "rename", "titles", etc. (and some good old-fashioned googling) but couldn't find anything. The history can sometimes be pieced together to some degree based on information in the ((Old moves)) template on the talk page, but those only cover moves that happened as a result of RMs (and sometimes they're missing RMs). Colin M (talk) 21:09, 3 June 2019 (UTC)

@Colin M: That's very interesting! Quite useless on the software's part. It would be nice if there was a page-move tag, which we could use in a page's revision history to find page moves and hence previous names. Sorry, I can't think of any immediate solution. Maybe a user script that gets 5000 edits at a time and filters out all summaries that do not include "moved page" or "moved to" (which I think is automatically added). GUYWAN ( t · c ) 00:03, 4 June 2019 (UTC)

This feature doesn't appear to exist, see phab:T152829 requesting it. — xaosflux ^Talk 00:23, 4 June 2019 (UTC)

Hi greetings, the user script named User:Kephir/gadgets/sagittarius (modified version User:Sam Sailor/Scripts/Sagittarius+) seems better (useful) to create and improve redirects. It is better to become a gadget. It may help to reach this to more users who wishes to create or maintain redirects. Gadgets will reach more users than a userscript can. There are many users trying to improve redirect pages with templates, etc. This will help them to do it easily. Hope yours shall consider this. Thank you.--PATH SLOPU 12:44, 4 June 2019 (UTC)

@Path slopu: is there an active maintainer you know of, because I'm not seeing one. Kephir appears to gone inactive years ago. — xaosflux ^Talk 14:21, 4 June 2019 (UTC)

@Xaosflux:Hi greetings, Sam Sailor is the Wikipedian who modified the script and created sagittarius+. He is active now. I think he can become a active maintainer. Regards.--PATH SLOPU 14:27, 4 June 2019 (UTC)

I'm no good at coding, but perhaps Enterprisey would take over my fork. What would the benefit be of turning the script into a gadget? Best, Sam Sailor 14:31, 4 June 2019 (UTC)

@Sam Sailor:It may help to reach this to more users who wishes to create or maintain redirects. Gadgets will reach more users than a userscript can. There are many users trying to improve redirect pages with templates, etc. This will help them to do it easily.--PATH SLOPU 14:34, 4 June 2019 (UTC)

Hello everyone, how to connect CSS page to JS page? I have a gadget (or a js page) and i want to change it's style, then created a css file with codes, but the style not effect to the gadget!

Note: I tried to move the code to mediawiki common.css and the style code would effect to the gadget. Thanks! --ئارام بکر (talk) 20:55, 4 June 2019 (UTC)

ئارام بکر if you are talking about a true integrated gadget, you can define the CSS along with the gadget on your definitions page(e.g. MediaWiki:Gadgets-definition). — xaosflux ^Talk 22:29, 4 June 2019 (UTC)

If this is about MoreMenu (the dropdown menus gadget we were discussing on my talk page), you can follow the instructions at meta:MoreMenu#Gadget installation. You will need to adjust the widths accordingly to match the width of the ckb translations. — MusikAnimal ^talk 01:48, 5 June 2019 (UTC)

@Xaosflux: Thank you very much! After we defined the CSS page on MediaWiki:Gadgets-definition, the gadget worked with CSS very well!

@MusikAnimal: this discussion is not about dropdown-menus gadget; It is about Discussion Closer gadget (i think so it is user script on this wiki, not a gadget). Thanks! --ئارام بکر (talk) 10:08, 5 June 2019 (UTC)

Does anyone know how to query the blockinfo of an IP address? I tried passing an IP to the ususers prop here: api/query/user, but, apparently, it doesn't work like that. It works fine with usernames, but I need to check whether an IP is blocked or not. Thanks. —RainFall 06:19, 5 June 2019 (UTC)

Use a query for blocks,[27] and don't forget global blocks.[28] So, try this:[29]. -- zzuuzz ^(talk) 06:49, 5 June 2019 (UTC)

@Zzuuzz: Oh, thanks. So, it's not possible to query a registered user and an IP with a single request—that's a bit weird design. Also, can I not pass multiple IPs separated with |? —RainFall 07:10, 5 June 2019 (UTC)

RainFall, This is correct. I have a lot of tools that need to pull lots of IP blocks at the same time. It's always one at a time, unfortunately. Documentation can be found at mw:API:Blocks. SQL^{Query me!} 08:16, 5 June 2019 (UTC)

You can use bkusers instead of bkip[30] but you will not get the info about indirect blocks. You can do a similar thing with global bgaddresses, however global accounts are locked instead of blocked, so that's a different check. -- zzuuzz ^(talk) 08:28, 5 June 2019 (UTC)

Looks like that is what I need. Thanks. —RainFall 11:17, 5 June 2019 (UTC)

File:Seal of Far Hills, New Jersey.jpg looks splendid, and stays sharp even when I zoom right out in my browser to make it very small. However, the thumbnail on the File: page and its appearance in Far Hills, New Jersey are badly blurred. Is there any way that the original artwork can be reproduced more faithfully at the smaller size? Thanks, Certes (talk) 13:44, 5 June 2019 (UTC)

No opinion on the reduction, but I really wonder if a seal can be CC-0 licensed. Jo-Jo Eumerus (talk, contributions) 14:03, 5 June 2019 (UTC)

Some government entities do release their work as PD (the US federal government being probably the best known), but this one only permits noncommercial use. I've flagged for deletion on Commons accordingly. Seraphimblade ^{Talk to me} 14:14, 5 June 2019 (UTC)

Sadly the documentation page has now gone, but I believe the uploader claimed to have created the artwork themselves rather than copying a non-PD image. Certes (talk) 14:57, 5 June 2019 (UTC)

Yes, but how can one create the seal of a political entity oneself? Unlike coats of arms, which are art concepts and not artworks by themselves - the organization describes how the coat should look like (e.g "a red cross with vertical and horizontal bars and green circles in each white space between the bars") but it's up to individual people to actually draw the coat - a seal is generally an image by itself. Jo-Jo Eumerus (talk, contributions) 15:03, 5 June 2019 (UTC)

If it's been deleted for not being the real seal rather than for any copyright reason, that makes sense. Certes (talk) 15:06, 5 June 2019 (UTC)

Subpages are not enabled in the main namespace but are in the talk namespace. This means that while, e.g. OS/2 is not a subpage of OS, Talk:OS/2 technically is a subpage of Talk:OS. Is there any way to suppress the display of that parent page link a the top of such talk pages? This does sometimes cause confusion, see for example Talk:Derry#subpage relating to Derry/Londonderry name dispute/Talk:Derry/Londonderry name dispute. If there isn't, is there any way there could be - maybe something along the lines of how ((wrong title)) works? Thryduulf (talk) 22:04, 4 June 2019 (UTC)

It's requested in phab:T41395: "behavior switch needed to disable links to parent pages from subpages". Talk:9 got some misplaced posts and now has a hatnote for users who got there by accident, but the hatnote has no way of knowing where the user came from. Special:PrefixIndex/Talk:9/ gives many possibilities. PrimeHunter (talk) 20:38, 5 June 2019 (UTC)

I just noticed that the "Global user contributions" tool that's linked to from the usual contribs page is giving no results (for myself, and a couple random people I tried). Sorry if this is the wrong place to raise this, or if it's already known. –Deacon Vorbis (carbon • videos) 02:09, 6 June 2019 (UTC)

Yes, it's not working. This has been reported on phab however. – Ammarpad (talk) 04:21, 6 June 2019 (UTC)