Wikipedia

Wikipedia:Village pump (technical)/Archive 184

< Wikipedia:Village pump (technical)
Village pump (technical) archive  

This page contains discussions that have been archived from Village pump (technical). Please do not edit the contents of this page. If you wish to revive any of these discussions, either start a new thread or use the talk page associated with that topic.

< Older discussions · Archives: A, B, C, D, E, F, G, H, I, J, K, L, M, N, O, P, Q, R, S, T, U, V, W, X, Y, Z, AA, AB, AC, AD, AE, AF, AG, AH, AI, AJ, AK, AL, AM, AN, AO, AP, AQ, AR, AS, AT, AU, AV, AW, AX · 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212

Expanding class=wikitable to make tables accessible to blind, and to align cell text

Tracked in Phabricator
Task T2418

Expanding what class=wikitable does could instantly solve a lot of problems the blind and visually impaired have with hundreds of thousands of tables on Wikipedia. And it would align row header cell text to the left. For more info see: User:Timeshifter/Sandbox114 and 115. See related discussion. (later note: link to archive: Wikipedia:Village pump (technical)/Archive 183#class="wikitable aligned linked" for linked country lists with flags).

Standard class=wikitable table below. It includes style=text-align:right; overall, and align=left on first-column cells.

Correctional supervision rates by state, 2016.




Jurisdiction 		Total Community supervision Incarcerated
Total,
12/31/2016 		Rate per
100,000
adults 		Probation
or Parole,
12/31/2016 		Rate per
100,000
adults 		In prison
or jail,
12/31/2016 		Rate per
100,000
adults
  Alabama 99,800 2,640 60,700 1,610 40,900 1,080
  Alaska 12,900 2,320 8,400 1,520 4,400 800
  Arizona 137,500 2,570 84,800 1,590 55,000 1,030

The blind need scopes on all header cells. Added row headers, scope=row, scope=rowgroup, scope=col, scope=colgroup, class=plainrowheaders:

Correctional supervision rates by state, 2016.




Jurisdiction 		Total Community supervision Incarcerated
Total,
12/31/2016 		Rate per
100,000
adults 		Probation
or Parole,
12/31/2016 		Rate per
100,000
adults 		In prison
or jail,
12/31/2016 		Rate per
100,000
adults
  Alabama 99,800 2,640 60,700 1,610 40,900 1,080
  Alaska 12,900 2,320 8,400 1,520 4,400 800
  Arizona 137,500 2,570 84,800 1,590 55,000 1,030

Same as above. Added style=background-color:yellow; and style=background-color:lightyellow; - I made one cell light yellow to show a couple shades of yellow.

Correctional supervision rates by state, 2016.




Jurisdiction 		Total Community supervision Incarcerated
Total,
12/31/2016 		Rate per
100,000
adults 		Probation
or Parole,
12/31/2016 		Rate per
100,000
adults 		In prison
or jail,
12/31/2016 		Rate per
100,000
adults
  Alabama 99,800 2,640 60,700 1,610 40,900 1,080
  Alaska 12,900 2,320 8,400 1,520 4,400 800
  Arizona 137,500 2,570 84,800 1,590 55,000 1,030

class=wikitable - It could be expanded to add scopes to all header cells, including those spanning columns and rows (colspan and rowspan). And it could align row header text to the left as does class=plainrowheaders. The scopes would allow users of screen readers (the blind) to understand nearly all tables on Wikipedia, since nearly all use class=wikitable.

class=wikitable could also add a yellow or light yellow background to headers. That would solve Aréat's problem with dark grey as a header background due to his visual impairment. It would help all readers. Highlighter felt-tip markers usually come in some shade of yellow. That is because this background color behind black text is one of the most legible color combinations.

class=plainrowheaders does not use bolded text in row headers. Many people do not like default bolded row headers. Especially in wide country lists, and other tables, where the width matters due to long country names, etc.. Also, people sometimes prefer to bold selected words or phrases in row headers.


Wikitext below is for the above yellow table if class=wikitable did scopes, color, and plain row headers. It's much simpler to edit by the average editor. Compressed horizontal format is more intuitive without all the inline CSS, etc.. 

{|class="wikitable sortable mw-datatable" style=text-align:right;
|+ Correctional supervision rates by state, 2016.
|- 
! rowspan=2 |<br><br><br><br>Jurisdiction
! colspan=2 |Total 
! colspan=2 |Community supervision 
! colspan=2 |Incarcerated
|-
! Total,<br>12/31/2016 
! Rate per<br>100,000<br>adults 
! Probation<br>or Parole,<br>12/31/2016 
! Rate per<br>100,000<br>adults 
! In prison<br>or jail,<br>12/31/2016 
! Rate per<br>100,000<br>adults
|-
! {{flaglist|Alabama}} 
|| 99,800 || 2,640 || 60,700 || 1,610 || 40,900 || 1,080
|-
! {{flaglist|Alaska}} 
|| 12,900 || 2,320 || 8,400 || 1,520 || 4,400 || 800
|-
! {{flaglist|Arizona}} 
|| 137,500 || 2,570 || 84,800 || 1,590 || 55,000 || 1,030
|}

Wikitext below is the actual wikitext for the above yellow table, and includes all the inline CSS, etc.: 

{|class="wikitable sortable mw-datatable plainrowheaders" style=text-align:right;
|+ Correctional supervision rates by state, 2016.
|- 
! scope=rowgroup rowspan=2 style=background-color:lightyellow; |<br><br><br><br>Jurisdiction
! scope=colgroup colspan=2 style=background-color:yellow; |Total 
! scope=colgroup colspan=2 style=background-color:yellow; |Community supervision 
! scope=colgroup colspan=2 style=background-color:yellow; |Incarcerated
|-
! scope=col style=background-color:yellow; |Total,<br>12/31/2016 
! scope=col style=background-color:yellow; |Rate per<br>100,000<br>adults 
! scope=col style=background-color:yellow; |Probation<br>or Parole,<br>12/31/2016 
! scope=col style=background-color:yellow; |Rate per<br>100,000<br>adults 
! scope=col style=background-color:yellow; |In prison<br>or jail,<br>12/31/2016 
! scope=col style=background-color:yellow; |Rate per<br>100,000<br>adults
|-
! scope=row style=background-color:yellow; |{{flaglist|Alabama}} 
|| 99,800 || 2,640 || 60,700 || 1,610 || 40,900 || 1,080
|-
! scope=row style=background-color:yellow; |{{flaglist|Alaska}} 
|| 12,900 || 2,320 || 8,400 || 1,520 || 4,400 || 800
|-
! scope=row style=background-color:yellow; |{{flaglist|Arizona}} 
|| 137,500 || 2,570 || 84,800 || 1,590 || 55,000 || 1,030
|}

--Timeshifter (talk) 19:13, 18 August 2020 (UTC)

Timeshifter, sorry, but i'm just really confused.. There is so much text and so much back and forth that went on, that i'm simply unsure what we are trying to achieve here... I think the idea is 'less inline styles'. Which I agree with. It seems you think you have solutions for multiple problems. It would benefit everyone to treat each problem separately to make sure that we have considered all upsides and downsides of any proposed solutions. But I think we can all agree that there is not a chance that we will ever see yellow headers as a default.
But lets start here:
  1. The blind need scopes on all header cells that is simply not true. While good advice in general, all screenreaders I have worked with simply make their own header direction analysis for tables (because no one other than Wikipedia uses the scope attribute) or even completely ignore the direction of headers. So in practice it isn't really a thing. There is a difference between being a 'need' and an 'improvement/net benefit'.They can also be very useful for styling.
  2. class=wikitable - It could be expanded to add scopes to all header cells I guess.. theoretically. By making the parser aware of wikitable class (it is only styling so far, not behaviour) and to dynamically analyse the generated HTML of the table and add scope attributes in a DOM post processing filter ? Is that what you are proposing ?
  3. class=wikitable could also add a yellow or light yellow background to headers. not gonna happen. We cannot adopt to every single persons impairment (i don't know where and when Areat stated his impairment, that would have helped). Primarily because adaptations for impairments can actually clash with the impairment of others (especially with colorblindness this is a problem). There is a balance to be struck. If someone is so much impaired that even while passing WCAG contrast levels, they still have a problem, then we can think about slightly readjusting the hue or contrast, but yellow backgrounds are way out there. Browsers have zoom levels, ppl can use high contrast mode of their OS and/or browser, install custom stylesheets in their user profile or even with stylish in their browser
  4. class=plainrowheaders does not use bolded text in row headers. I'm not sure what the proposal here is. You want to make plainrowheaders the default for row headers with scope=row/rowgroup ? Have we checked which tables depend on this behavior currently (like tables without column headers ?)
  5. You are also mixing wikitable and mw-datatable I noticed.. not sure what you want there.
I hope I am making sense. —TheDJ (talk • contribs) 10:01, 19 August 2020 (UTC)
TheDJ, Thanks for replying. Are you visually impaired? You wrote that you use a screen reader. I would like to run some sandbox tests with you concerning tables with and without scope=rowgroup and scope=colgroup.
  1. no one other than Wikipedia uses the scope attribute. Apparently, WCAG no longer requires using scopes on simple tables. It is unclear to me if a simple table includes tables with rowspan and colspan. If rowspan and colspan are still a problem for the blind, then maybe class=wikitable could add scope=rowgroup whenever it sees rowspan, and scope=colgroup whenever it sees colspan. I would think that would be fairly simple.
  2. class=plainrowheaders does not work without scope=row. So if scope=row is not needed for simple tables, then there is also no need for class=plainrowheaders in class=wikitable.
  3. class=mw-datatable is just in the table by habit of mine. It does not need to be in class=wikitable, though it would be nice because it is really useful for helping people follow a line across a table. Now that I think about it more, that is also something that helps the visually impaired even more. So maybe it should be in class=wikitable.
  4. Browsers have zoom levels, ppl can use high contrast mode of their OS and/or browser. I don't think the visually impaired should have to adjust those levels constantly depending on whether a Wikipedia table has a column with a darker gray background or not. That is Aréat's problem, I believe, from reading his user talk page. Hopefully, he will explain it further here. A light yellow background is a lot better for everybody. Wikipedia has a mix of tables with and without rowheader columns (with the darker gray background). Right now I have my brightness turned down due to getting up from sleeping. That makes the background of the whole page light gray. In this case the additional gray darkening in header columns is a problem even for me. And I am not visually impaired other than needing glasses. But when I look at the above tables I have no problem with the header cells with a yellow background.
  5. From what I read the colorblind would see light yellow as light pink. So for them they still see some header cell background shading. Also, yellow colorblindness is much less common than for other colors. See color blindness and this page. So I support a light yellow background for header cells. A light pastel yellow color versus a strong yellow color.
  6. style=text-align:left; should be the default on rowheader columns. That is what most tables need. The relatively few tables that don't want it can add style=text-align:center; or style=text-align:right; to each rowheader cell. By the way I discovered that align=left does not work on rowheader cells on Wikipedia. It only works on data cells.
  7. I think the idea is 'less inline styles'. Which I agree with. All of the above, other than light yellow backgrounds, would remove some clutter from table wikitext, and make table editing easier even for newbies.
Here are some relevant excerpts from Aréat's talk page. From looking at his user page I think Aréat may be a native speaker of French. So that may explain the confusion with some of his messages there and elsewhere. From Aréat's talk page:
I've read it, and seen nothing indicating the left column must be centered, bolded on dark grey background. It make it less easier to read in the first place. This is absurd. ...

Look at the table you're proposing and honestly tell me it's easier to read a text that is black on dark grey? This [MOS] is obviously wrote by people who don't use it. You can't make it less easier to read on a whim. ...

I have a very poor vision and I reverted the changes you made to a table that made it way more difficult to read, and that isn't used anywhere on the election pages I've been participating on for years, where the tables have been fine to read. I reverted the pages to the version they were before you went there, imposed your changes and kept reverting despite being asked to go to the talk page.

I really dislike that solid gray background of the above table using class=wikitable
Below is part of that Aréat info, but with a light yellow background. Turn down the brightness to 60% and 40% to see the difference even more. I use freeware CareUEyes Lite. It changes brightness levels instantly from a clicker icon on the system tray in Windows.
Look at the table you're proposing and honestly tell me it's easier to read a text that is black on dark grey? This [MOS] is obviously wrote by people who don't use it. You can't make it less easier to read on a whim. ... I have a very poor vision
--Timeshifter (talk) 17:47, 19 August 2020 (UTC)
Timeshifter, Are you visually impaired? You wrote that you use a screen reader. no, but as a developer I have an interest in accessibility and regularly test Wikipedia and esp. new features with VoiceOver and NVDA and then slap WMF for having written inaccessible widgets again. I'll get back on the other points when I have time. —TheDJ (talk • contribs) 08:35, 20 August 2020 (UTC)
  1. if a simple table includes tables with rowspan and colspan Simple tables generally have only one level of headers for columns and/or one level of headers on the rows.
  2. class=plainrowheaders does not work without scope=row well.. how would you separate the simple table from the 'non simple' table though ? CSS cannot determine this based on just a TH, because the TH can be everywhere in the table....
  3. "it is really useful for helping people follow a line across a table" I don't mind row highlighting, but 1, what if it is not a datatable and 2. i think we shouldn definetly change the background color of TD cells inside wikitables in that case.
  4. I don't think the visually impaired should have to adjust those levels constantly 1. why would that have to be constantly ? Set it and its done 2. I do think that our responsibility on accessibility stops somewhere. I'm pretty sure I can find evidence that a page with white background and black letters with a 24pt font size might be the most readable version of a website to a large portion of the populace. But there are also people who get blinded by white and most of us actually appreciate a bit of design esthetic. Everything is a balance. Lastly I will also point out that some ppl tend to use (oft not by choice) ridiculously cheap computer equipment which might be their bigger problem. If you have a 300 euro laptop, its like wearing 15 euro plastic glasses. You are better off getting 2nd hand, older but higher quality, equipment sometimes.
  5. I won't argue that it isn't a good color for accessibility reasons. I just think too many people would hate it for it to be a realistic candidate.
  6. should be the default on rowheader columns I agree.. but how to accurately detect row headers when some tables are complex ? I mean, we could ADD a rule that left aligns rows headers on simple tables, but that means that on 'non-simple' tables, you would get left aligned text in the first and middle aligned in the second, and in some ways, less predicatble behavior like that is worse than having a consistently working solution with "plainrowheaders". Instead, why not just make the current plainrowheaders (with scope) the actual default ? align=left that's because it is not supported in HTML5 any longer and therefore the default styling of our headers (center aligned) gets priority over any "align" instruction.
All in all, I agree with many of your points, but I think that with the diversity of Wikipedia's often highly complex tables and the limited styling options that HTML really gives to safely support those tables in a generic instead of a case by case way, you are overestimating the effectiveness of some of those proposals. Tables on Wikipedia are HARD, much harder than anywhere else on the web I would postulate. I'm also personally in favor of toning down the default background color of both TD and TH cells down considerably for accessibility reasons, but as one of the oldest color combinations of English Wikipedia (2006ish ?), I think that would require a separate discussion, with multiple variants for the community to vote on. —TheDJ (talk • contribs) 14:37, 23 August 2020 (UTC)
  1. Thanks for replying, TheDJ. I think I am asking for too much change at once of class=wikitable. This much change needs paid staff working on getting consensus. Are you paid staff? I think it may be better to do smaller changes of class-wikitable. A step at a time. Otherwise nothing may get done.
  2. Thanks for the link to the WCAG page with a clearer description of what a simple table is.
  3. Do your screen readers recognize headers that span multiple columns (colspan)? Do they recognize headers that span multiple rows (rowspan)?
  4. About row highlighting. I don't want permanent row highlighting. I want it like class=mw-datatable does it. Only when a cursor is over the row. As for the background color of the row, I want it to be a very light pastel color. Anything but gray.
  5. Left-aligning of text in the first column is necessary in most tables regardless of whether that column consists of headers or not. A table template has been created (not by me) just to be able to quickly do daily updates of COVID-19 pandemic death rates by country. In order to have the first column text left-aligned. It uses this template stylesheet. Imagine if this happens on thousands of tables.
  6. I think it is much easier to add left-alignment of the first column to class=wikitable. Of course then you would need something like data-sort-type except for text alignment. For those few first columns that don't need left alignment.
  7. I wonder if all the class=wikitable CSS/JS could be placed in separate CSS and JS files that are uploaded to the browser only when class=wikitable is seen by the Mediawiki software. That would lower the CSS/JS loads to the Wikipedia articles without tables.
  8. Or better yet would be separate separate CSS and JS files for any Wikipedia article containing signs of any kind of table: Via the mediawiki software spotting {| or <table>.
--Timeshifter (talk) 07:15, 24 August 2020 (UTC)

Comment. TheDJ, I see from your user page that you are not paid staff, and are a volunteer like me. It looks like major changes are not going to happen. At least not right away. I would be happy if something simpler like this below would happen. But instead of forcing a sorting type for a column it would set the text alignment for a column:

It could be added to any column header: data-align="..." Text alignment in the column could be set for left, center, or right. Related discussions:

An example of a table with varying needs for text alignment is the main country table in this article:

  • List of countries by intentional homicide rate - the country, region, and subregion columns need their text aligned left. The rate and count columns need their data aligned right. So there are 3 text columns aligned left, and 2 data columns aligned right. Inline CSS is currently used to align every data cell to the right. That is a lot of clutter, and unnecessary work.

I guess the next logical step is to propose this at Wikimedia Phabricator. --Timeshifter (talk) 02:40, 28 August 2020 (UTC)

Timeshifter, There is phabricator T2418 which may be of interest. -- WOSlinker (talk) 06:41, 28 August 2020 (UTC)
WOSlinker, thanks! I just subscribed to that task. I will post some ideas there. Hopefully soon. --Timeshifter (talk) 07:20, 28 August 2020 (UTC)

Image positioning changes on mobile?

I recently noticed that on mobile, an image always appears after the first text paragraph even if it's located at the top of the page (in Wikitext). This behaviour was previously only seen with infoboxes. Why was this change made? If people wanted the image to appear after the paragraph, they would simply add it like that in Wikitext. Forcing image positioning is unnecessary. Examples of affected articles: Jai Shri Ram, Ramnami Samaj. Regards, TryKid [dubious – discuss] 12:22, 28 August 2020 (UTC)

On Jai Shri Ram, the {{Confuse}} template, which appears in the wikitext after the first image but before the lead paragraph text, is rendered below the lead paragraph on mobile. That seems like a bug. I realize that the Confuse template may not be exactly in the right place, but it should render above the lead paragraph. – Jonesey95 (talk) 16:03, 28 August 2020 (UTC)

Basic Lua help needed at Template talk:Bibleverse

Someone with pretty basic Lua skills is needed at Template talk:Bibleverse. Thanks. – Jonesey95 (talk) 17:18, 28 August 2020 (UTC)

Twinkle, RedWarn not showing up

Moved from WP:Teahouse. Ed talk! 20:18, 28 August 2020 (UTC)

I have Twinkle and RedWarn enabled, but they're not showing up for some reason. I think it started today or last night. Mvcg66b3r (talk) 19:45, 28 August 2020 (UTC)

Hi, @Mvcg66b3r can you open your Javascript console and check for errors? See Wikipedia:Reporting JavaScript errors Ed talk! 19:48, 28 August 2020 (UTC)
I'm using Chrome 85 with Windows 10 version 2004. This is the error message that came up:

Uncaught (in promise) Error: The message port closed before a response was received. 

   at document_start.js:8

Mvcg66b3r (talk) 20:12, 28 August 2020 (UTC)

gutenberg: links broken

The English Wikipedia appears to have a special prefix, [[gutenberg:, to link to Project Gutenberg books. It is currently broken, apparently because PG has changed their URL structure from /etext/ to /ebooks/ without leaving a site-wide redirect. I'm sure that Michael Hart would have been appalled at this basic misstep in web practice, but here we are.

Does anyone know how to change the URL that is used by [[gutenberg:? I searched all namespaces at en.WP and at mediawiki.org and was unable to find anything that explained how this magical prefix works.

Please see this template talk discussion for the original thread. – Jonesey95 (talk) 16:28, 28 August 2020 (UTC)

@Jonesey95: It's at meta:Interwiki map and changes are requested at its talk. Johnuniq (talk) 23:33, 28 August 2020 (UTC)
Thanks. I left a note there. – Jonesey95 (talk) 02:34, 29 August 2020 (UTC)

Can we flag broken links to YouTube?

See Template_talk:YouTube#Broken_links_need_to_be_flagged where I noted the problem a few days ago (but I guess that is a low-visibility talk page). --Piotr Konieczny aka Prokonsul Piotrus| reply here 01:25, 29 August 2020 (UTC)

@Piotrus: not automatically, there is nothing special about youtube, it would be the same as any other broken web link. Some of the reference archive bots might look in to it - but really what would you want done with the link? — xaosflux Talk 02:02, 29 August 2020 (UTC)
The archive bots would not detect as dead because it returns 200 ie. a soft-404.
./header 'https://www.youtube.com/watch?v=QkWIN7SXowc'
HTTP/1.1 200 OK
It would require a specialized bot to detect. But even then, sites like archive.org don't save youtube videoes (with a few exceptions). Typically dead youtube videos are gone forever, which is to say, all in time will evaporate. -- GreenC 02:38, 29 August 2020 (UTC)
What I wold like to be done is for the dead YouTube link to be either marked as such (so it can be manually saved, since videos are sometimes reuploaded - like in the case of some public domain movies that first caught my attention) - or just removed. The first option is preferable as we need a human to decide whether a link can be rescued (public domain etc.) or is just gone. --Piotr Konieczny aka Prokonsul Piotrus| reply here 03:09, 29 August 2020 (UTC)
The web archive sites often have snapshots of everything in the page but the video. Thus when they get tagged with a {{dead}} the bots would add such an archive giving the false appearance of a "saved" video. For example from the article Google : https://web.archive.org/web/20100816150501/http://www.youtube.com//watch?v=soYKFWqVVzg -- GreenC 13:46, 29 August 2020 (UTC)

Mobile screen width set to iPad width?

Up until a few days ago my iPhone view was automatically set to the correct width (I almost exclusively use it vertically).

Now frustratingly the screen width seems to be auto set for a tablet width so no view of mine shows the full text on the screen.

Help! Gleeanon409 (talk) 02:40, 30 August 2020 (UTC)

Help with gerrit and LinksUpdateHookHandler.php

Hey all - I'm trying to document some of the workings of LinksUpdateHookHandler.php in a user space essay and I don't know gerrit well enough to know if I'm looking at the most current version. I'm trying to find the equivalent of what used to be here https://github.com/wikimedia/mediawiki-extensions-PageImages/blob/master/includes/LinksUpdateHookHandler.php

Can anyone help? And/or could you point me to some information to get me up to speed on gerrit? - Scarpy (talk) 05:31, 30 August 2020 (UTC)

Scarpy, I personally browse on the Phabricator mirror ( https://phabricator.wikimedia.org/diffusion/EPIM/ particularly) since I know it syncs close to daily (I do not know the Github cadence) and the search interface is simple enough for me to navigate. I think from a browsing perspective on gerrit you want https://gerrit.wikimedia.org/g/mediawiki/extensions/PageImages (you can find it from Browse in the new GUI with some effort).
The file in question is at includes/Hooks/.. rather than at includes/.. (per PSR 4 move I believe). I would guess Github is reasonably up to date since I can find the file in the same directory in all three repo/mirrors. --Izno (talk) 06:08, 30 August 2020 (UTC)

Article draft submitted as jpg

I discovered this doozy today: File:Rogemar Ecaranum.jpg. Somehow, this was submitted as a jpg, instead of Draft:Rogemar Ecaranum. I can't move it to the correct place. Not sure what to do, or how this was even possible.--Auric talk 10:16, 30 August 2020 (UTC)

I mean, we should know what to do for the future, but in this case I'm rather inclined to just delete it and let the author know where to attempt to recreate it. Ian.thomson (talk) 10:28, 30 August 2020 (UTC)
Thanks. Tagged for CSD under F2, F10 and G8.--Auric talk 10:45, 30 August 2020 (UTC)

Charinsert problem

When writing User:Jo-Jo Eumerus/Gemmi Pass Fault I've noticed that the characters below the edit box cannot be inserted. I click on them and nothing happens. Jo-Jo Eumerus (talk) 11:59, 28 August 2020 (UTC)

That's mw:Extension:CharInsert (MediaWiki:Gadget-charinsert-core.js) we're talking about. ​ Which of those characters aren't working? All the ones I checked worked for me. SD0001 (talk) 12:48, 28 August 2020 (UTC)
None of these that I checked works. BTW, I think it's a Charinsert thing b/c shutting it off in Preferences causes all the buttons to disappear (and reappear upon switching it back on). Jo-Jo Eumerus (talk) 13:46, 28 August 2020 (UTC)

Jo-Jo Eumerus: You are importing quite a few user scripts in User:Jo-Jo Eumerus/common.js, some of them old or otherwise causing script errors. For example: if I enter

importScript( 'User:Ais523/adminrights.js' );

I get

You installed the userscript [[User:Ais523/adminrights.js]]
It has been deprecated and replaced with [[User:Amalthea/userhighlighter.js]]

and

importScript('User:Ohconfucius/script/MOSNUM dates.js');

yields Uncaught ReferenceError: ohc_US_slash_dates_driver is not defined. Nirmos (talk) 23:07, 29 August 2020 (UTC)

Ha. @Nirmos:, removing these scripts did in fact resolve the issue. However I wonder where you see the "ohc_US_slash_dates_driver is not defined" issue as I can't see it. Jo-Jo Eumerus (talk) 07:41, 30 August 2020 (UTC)
Jo-Jo Eumerus: In the browser console. You should be able to open the console in most browsers (Chrome/Edge/Firefox) by pressing f12. Nirmos (talk) 08:26, 30 August 2020 (UTC)
Well, F12 does open the console but not that warning. I suspect that it's not clear which console on which URL. Perhaps someone who sees the warning can file a request at User talk:Ohconfucius/script/MOSNUM dates to fix the problem. Jo-Jo Eumerus (talk) 09:16, 30 August 2020 (UTC)
@Jo-Jo Eumerus: You also have the discontinued User:AWeenieMan/furme.js installed in your vector.js. It or some other script not mentioned by Nirmos may be interfering with Charinsert. Nardog (talk) 23:04, 30 August 2020 (UTC)

Free Image Search Tool error

As I posted in Template talk:WikiProject Biography#Free Image Search Tool error a month ago, the link to the Free Image Search Tool generated by {{WikiProject Biography}} generates a 404 error. Is this something that can be fixed? Thanks! GoingBatty (talk) 05:01, 30 August 2020 (UTC)

GoingBatty, you will need to contact the tool maintainer, who is Magnus Manske. --Izno (talk) 05:45, 30 August 2020 (UTC)
@Izno: Thanks - I opened an issue for Magnus Manske at https://bitbucket.org/magnusmanske/fist/issues/15/free-image-search-tool-link-generates-a GoingBatty (talk) 00:44, 31 August 2020 (UTC)
Oops, I've already fixed the link in the template. But the current location of the tool is quite redundant and most likely not intended (which is a result of the recent move of Toolforge tools to subdomains), so it should probably be fixed anyway. Nardog (talk) 00:52, 31 August 2020 (UTC)
Nardog, yeah, I didn't think it necessary to call out the issue with the link since I tried with the good link and found the service did not work. --Izno (talk) 01:05, 31 August 2020 (UTC)
GoingBatty, conisder emailing him as well. I don't know how much he checks BB, even though that is the ostensible place to bug him. --Izno (talk) 01:05, 31 August 2020 (UTC)

Intersection Observer alternative

Tracked in Phabricator
Task T261609

Hi!

Wikipedia recently changed the way lazy loading is implemented. It now makes use of the Intersection Observer API.

Many devices (for example all iOS devices below 12.2) do not support this API, which renders wikipedia almost unusable to read certain types of articles, particularly math related, since the fallback solution has the user tap a tiny tiny grey box for every image in the article. When it comes to math a single character in a sentence might be an image.

This is quite sad, since it changes wikipedia from the fastest and most reliable source of information, to a website you only visit if you don’t have any other options.

Is there any way to make wikipedia usable again on older devices? Maybe a way to disable lazy loading altogether, or make wikipedia use the old algorithm which actually worked before falling back to the grey box solution. — Preceding unsigned comment added by Kilian Geronimo (talk • contribs) 15:55, 29 August 2020 (UTC)

Kilian Geronimo, I have filed: T261609 —TheDJ (talk • contribs) 08:50, 31 August 2020 (UTC)

Mobile diff overflow / scrolling

Recently (I first noticed it yesterday) diffs on mobile containing URLs don't add line breaks and instead just let the content scroll out to the side, making diffs really annoying. Is this a recent accidental change?  Nixinova   T   C   01:21, 29 August 2020 (UTC)

Nixinova, this is likely related to an effort where they are trying to bring the mobile diff and the desktop diff together again in the same view. I've made note of this problem (which is likely unintended) in the ticket phab:T240622. —TheDJ (talk • contribs) 08:59, 31 August 2020 (UTC)

I think this new “feature” is what’s bothering me as well. In mobile view, which I use almost all the time, I used to have all the editing screens justified (with no need to scroll sideways) while using my phone upright. That has changed for the worse in the last few days. Gleeanon409 (talk) 15:35, 29 August 2020 (UTC)

Halp with ugly very long URL

There is an eye sore at [1]. My wiki-fu is not strong enough to make it bearable, can you help? --Palosirkka (talk) 20:45, 24 August 2020 (UTC)

While this is more of a new problem (unless you have the same problem) than a solution, I see this URL all the way across the page, covering up information in different columns. It doesn't even happen in every line of the URL.— Vchimpanzee • talk • contributions • 20:50, 24 August 2020 (UTC)
I also see that, probably should've mentioned it. --Palosirkka (talk) 08:10, 25 August 2020 (UTC)
@Palosirkka and Vchimpanzee: I have hidden the stray note in an HTML comment: Special:Diff/974837864. --CiaPan (talk) 09:14, 25 August 2020 (UTC)
Thank you CiaPan, I think that was a good solution as the info does not need to be visible in the refs. --Palosirkka (talk) 09:26, 25 August 2020 (UTC)
@Palosirkka: Thank you for your comment. I wonder if the part I hid should remain there at all, or rather just be deleted. When hidden, it does not help our readers. It's not very useful when viewed with a text editor, either. If anybody wants to configure or parametrize the linked report, they can achieve that with controls available there in. Geek hints on URL hacking are quite outside the scope of encyclopedia... I think I'll delete that in a few days, unless someone tries to restore it or utilize in other way. --CiaPan (talk) 11:31, 25 August 2020 (UTC)
@CiaPan:I tried to take a look at the URL but it outputs nothing without JavaScript so it's impossible for me to say whether that snippet is of use or not. --Palosirkka (talk) 11:33, 26 August 2020 (UTC)
Deleted: Special:Diff/975951798. --CiaPan (talk) 10:09, 31 August 2020 (UTC)
Notifying @Palosirkka and Vchimpanzee:.

Tool for running example numbering + automated referring-back

Hi! I was wondering if it would be possible to develop a tool which allows (1) numbered lists whose numbering continues throughout the article and (2) an automated way of referring back to earlier examples. I would find this enormously useful for linguistics articles, and another editor at the help desk suggested that this might be useful for many other math/science related articles as well. Speaking just for linguistics articles, many of them currently format their examples in ugly or confusing ways, and have to use text that “breaks the fourth wall” in order to achieve a minimum standard of clarity. My idea is that editors could write markup like this (or perhaps less cumbersome!):

Example <exref = “donkey” /> shows that donkey anaphora exists. Example <ex ref = “donkeyRC”> uses a relative clause, while Example <ex ref = “donkeyif /> uses a conditional.
<ex name=“donkey” /> Donkey Anaphora:
<subex name = “donkeyRC” />Every farmer who owns a donkey beats it.</subex>
<subex name = “donkeyif” />If a farmer owns a donkey, he beats it.</subex>
</ex>
Blah blah blah imagine this text runs for several paragraphs, perhaps with a section break in the middle. Now compare the pair shown in <ex ref = “donkey” /> to the completely unrelated ones in <exref = “kangaroo” />.
<ex name=“kangaroo” /> Counterexample to Antecedent Strengthening:
<subex name = “kangarootail” />If Kangaroos didn't have tails, they would topple over.</subex>
<subex name = “kangaroojetpack” />If Kangaroos didn't have tails, but did have jetpacks, they wouldn't topple over.</subex>
</ex>

And this would render as:

Example (1) shows that donkey anaphora exists. Example (1a) uses a relative clause, while Example (1b) uses a conditional.
(1) Donkey Anaphora:
a. Every farmer who owns a donkey beats it.
b. If a farmer owns a donkey, he beats it.
Blah blah blah imagine this text runs for several paragraphs, perhaps with a section break in the middle. Now compare the pair shown in (1) to the completely unrelated ones in (2).
(2) Counterexample to Antecedent Strengthening:
a. If Kangaroos didn't have tails, they would topple over.
b. If Kangaroos didn't have tails but did have jetpacks, they wouldn't topple over

Crucially, the numbers would all update consistently if the page is edited to change the order of the examples, or to stick in intervening examples. I imagine that something like this could be built based on whatever currently handles reference numbering, and if any further inspiration is needed, the LaTeX package “linguex” has this functionality. Botterweg14 (talk) 16:51, 31 August 2020 (UTC)

Editing news 2020 #4

Read this in another language • Subscription list for this newsletter

Reply tool

 
The number of comments posted with the Reply Tool from March through June 2020. People used the Reply Tool to post over 7,400 comments with the tool.

The Reply tool has been available as a Beta Feature at the Arabic, Dutch, French and Hungarian Wikipedias since 31 March 2020. The first analysis showed positive results.

  • More than 300 editors used the Reply tool at these four Wikipedias. They posted more than 7,400 replies during the study period.
  • Of the people who posted a comment with the Reply tool, about 70% of them used the tool multiple times. About 60% of them used it on multiple days.
  • Comments from Wikipedia editors are positive. One said, أعتقد أن الأداة تقدم فائدة ملحوظة؛ فهي تختصر الوقت لتقديم رد بدلًا من التنقل بالفأرة إلى وصلة تعديل القسم أو الصفحة، التي تكون بعيدة عن التعليق الأخير في الغالب، ويصل المساهم لصندوق التعديل بسرعة باستخدام الأداة. ("I think the tool has a significant impact; it saves time to reply while the classic way is to move with a mouse to the Edit link to edit the section or the page which is generally far away from the comment. And the user reaches to the edit box so quickly to use the Reply tool.")[2]

The Editing team released the Reply tool as a Beta Feature at eight other Wikipedias in early August. Those Wikipedias are in the Chinese, Czech, Georgian, Serbian, Sorani Kurdish, Swedish, Catalan, and Korean languages. If you would like to use the Reply tool at your wiki, please tell User talk:Whatamidoing (WMF).

The Reply tool is still in active development. Per request from the Dutch Wikipedia and other editors, you will be able to customize the edit summary. (The default edit summary is "Reply".) A "ping" feature is available in the Reply tool's visual editing mode. This feature searches for usernames. Per request from the Arabic Wikipedia, each wiki will be able to set its own preferred symbol for pinging editors. Per request from editors at the Japanese and Hungarian Wikipedias, each wiki can define a preferred signature prefix in the page MediaWiki:Discussiontools-signature-prefix. For example, some languages omit spaces before signatures. Other communities want to add a dash or a non-breaking space.

New requirements for user signatures

  • The new requirements for custom user signatures began on 6 July 2020. If you try to create a custom signature that does not meet the requirements, you will get an error message.
  • Existing custom signatures that do not meet the new requirements will be unaffected temporarily. Eventually, all custom signatures will need to meet the new requirements. You can check your signature and see lists of active editors whose custom signatures need to be corrected. Volunteers have been contacting editors who need to change their custom signatures. If you need to change your custom signature, then please read the help page.

Next: New discussion tool

Next, the team will be working on a tool for quickly and easily starting a new discussion section to a talk page. To follow the development of this new tool, please put the New Discussion Tool project page on your watchlist.

Whatamidoing (WMF) (talk) 18:47, 31 August 2020 (UTC)

Layout problem

In the Bengali language article all the images in the top section have, for some reason, pushed downwards. I have tried fixing it to no success, have posted to the Teahouse, with no success (here and here). May be I am phrasing the request for help inadequately.

If you take a look at the article, I tried using one image with History, another image with the following section, and a third image in the next section, followed by a multiple image template in the next (all aligned to left, as the right side is taken over by infoboxes). But all of them has been pushed down, and now they are all bunched around toward the bottom end of {{Bengalis}} infobox.

I tried a similar thing in another article – putting one image in the first section. That one was pushed down a bit too. I was suggested that too many floating objects may be the problem, but the second article has only two floating objects at the top of the article – one image and one infobox – while the remedy suggested by Extended image syntax page didn't work (i.e. "alternating between left and right in aligning images or other floating objects").

How can I get the images at Bengali language article to appear where they were placed? Aditya(talk • contribs) 06:44, 30 August 2020 (UTC)

  Ahem. Did I, in my foolishness, come to a wrong page or posted a wrong question? I really thought I would get a solution either at Teahouse or here. Well... my bad. :( Aditya(talk • contribs) 21:10, 30 August 2020 (UTC)
Aditya Kabir I don't know if I exactly understood, or helped. Let me say that the image spacing looks entirely different in my Firefox browser, than it does in Chrome. So, that might be one factor. Also, sometimes I see where an image has pushed down into another section, because there is not enough section wording to even it out. In this case, I inserted the template {{clear}} after the first three images. If that resulted in what you wanted, then good. If it didn't, just revert my edits. That's the only thing I can think of to do. — Maile (talk) 22:53, 30 August 2020 (UTC)
This is a common problem. What you have here is three floating objects in the lead section - {{Infobox language|...}}, {{Contains special characters|Bengali}} and {{Bengalis}} each of which is right-aligned, and naturally the second and third are pushed down by those above. Later on (in the History section) you have a fourth floating object, File:East-magadhan-proto-languages.png. Even though this is left-aligned, its upper edge cannot be any higher than the upper edge of the previous floating object, the {{Bengalis}}. If you want the image to be displayed higher up than {{Bengalis}}, the latter needs to be moved to some point after the image. --Redrose64 🌹 (talk) 23:17, 30 August 2020 (UTC)
@Redrose64: Thanks. I get it now. This is what I was afraid of. Multiple infoboxes on the right makes it impossible to put an image in the top section across Wikipedia. But, even with Bengali Language forgotten, this looks like a problem for way too many articles (unless editors decide to give up and put no image in the top sections). Isn't that so? Aditya(talk • contribs) 02:03, 31 August 2020 (UTC)
It's a limitation of CSS, not of Wikipedia. --Redrose64 🌹 (talk) 19:19, 31 August 2020 (UTC)

Tech News: 2020-36

Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.

Problems

  • This is a reminder. All wikis will be read-only for a few minutes on September 1. This is planned between 14:00 and 15:00 UTC. Please check on the details on the announcement message. [3][4]

Changes later this week

  • There is no new MediaWiki version this week.

Tech news prepared by Tech News writers and posted by bot • Contribute • Translate • Get help • Give feedback • Subscribe or unsubscribe.

20:08, 31 August 2020 (UTC)

Move history tool

Is there a tool that allows you to see all moves a page has gone through? Special:Log is useless because it only lists moves when you enter the old name (if only it accepted a page ID...). Nardog (talk) 01:11, 31 August 2020 (UTC)

The logs only exist for the "old" page names. So, assuming no other loggable actions such as protection, if a page starts of at A and is moved to B and later on moved to C, the log for A will show a create and one move (to B); the log for B will show one move (to C); the log for C will be empty. To track a chain of moves it's usually necessary to look at the edit history of the page. --Redrose64 🌹 (talk) 19:25, 31 August 2020 (UTC)
@Redrose64: I know, that's exactly what I meant and I'm specifically asking for a way to automate that process. Disentangling the move history of a page can be quite a hassle especially when the page has a large number of revisions or when the same name has been occupied by multiple pages (e.g. "Cantonese"). Nardog (talk) 20:34, 31 August 2020 (UTC)
Yeah, it would be really nice if there were a way to log these. The nearest thing I know is the way previous RM discussions are normally tracked in the {{old RM}} banner on talk pages (but this practice hasn't been consistent over time). One way of getting the move history of a given page (not title), is to load its history (all of its history, possibly altering the &limit= bit in the url to move past the 500 limit), and then searching for the string moved (note the spaces on either side), which should fish out the moves by their automated edit summaries (this doesn't help with moves from the earlier days of Wikipedia when there were no such summaries). – Uanfala (talk) 15:43, 1 September 2020 (UTC)

Stop forcing readers to use the ugly interface

Hello, I am a reader of the English Wikipedia who is angry because I have to click each time to switch the view to the "mobile view", that is the only view that is modern, and readable by current standards. However, I feel like the English Wikipedia community is disrespectful towards us readers, because we do not participate in the Bizantine discussions about the sex of the angels. I do not care which interface you choose when you are logged in, and you have made it to your 50th year as an editor with your x286. For us contemporary people, there is no doubt that what you call the "mobile interface" must be the default one. I am horribly upset and disgusted by the paternalistic attitude of older Wikipedia editors that want to force us, readers, to follow their interface preference. Please change that attitude, and change the default view to the mobile view for all non-logged-in users. Logged in users can keep using their preferred interface as usual.--DataAthlete (talk) 21:58, 26 August 2020 (UTC)

Go to your preferences and set your skin to "Timeless" and be free.--Jorm (talk) 22:01, 26 August 2020 (UTC)
Wikipedia continues to have separate mobile and desktop sites. Why?.--Moxy 🍁 22:04, 26 August 2020 (UTC)
Timeless is a very nice skin for reading; alternatively you may be interested in opting in to and following the improvements to the Vector skin. There is certainly a discussion to be had about whether editors' views have a disproportionate impact on aspects of the interface, but it is beyond the scope of this page. – Teratix ₵ 04:37, 27 August 2020 (UTC)
@Jorm, Moxy, and Teratix: Your answers are a reflection of what is becoming Wikipedia, a useless site run by people who do not understand basic requests. I have been very clear in my request, but since it did not come across the first time, I will try again. I want English Wikipedia editors to stop forcing English Wikipedia readers to use the ugly interface. For this reason I want the "mobile interface" to become the default interface for all non-logged in users. How do I make that happen?--DataAthlete (talk) 07:30, 27 August 2020 (UTC)
Being rude and unpleasant to volunteers who have answered your questions is unlikely to persuade them to follow up with clarifications. – Teratix ₵ 08:15, 27 August 2020 (UTC)
The problem with your suggestion is that it effectively forces non-logged in users to use the mobile interface - where the reader doesn't see things like infoboxes, navboxes etc, so they don't get the same information.Nigel Ish (talk) 08:20, 27 August 2020 (UTC)
@DataAthlete: You can't fix it for non-logged in users. You can only affect your own destiny. I tried for years to make this shit-interface better for everyone and it is simply not fucking possible to do. Save your energy and make your own life better. The Wikipedia I see is pleasant and usable. I have my own CSS rules that fix lots of ugly. Go with whatever deity pleases you.--Jorm (talk) 15:34, 27 August 2020 (UTC)
  • @DataAthlete: - I am also concerned by the paternalistic attitude of certain Wikipedia readers that want to force everyone else, to follow their interface preferences, despite it concealing significant critical information, like infoboxes. I'm also at a loss on why your opening message (let alone your second) was so randomly hostile and rude, and also why you might think that is a good way to get the result you desire. Nosebagbear (talk) 08:27, 27 August 2020 (UTC)
Hi, DataAthlete. Could you, please, specify, how many 'contemporary people' are you, exactly? I think it would add appropriate weight to your request when other users see how many people you are. --CiaPan (talk) 09:51, 27 August 2020 (UTC)
  • Wikipedia exists for its readers, but would not exist without editors, and our main way of getting new editors is for readers to click the edit button and start editing. So of necessity the skins we show readers are a compromise between something tuned for readers and something designed for editors. Currently the mobile skin is much more reader focused than the desktop skin, and as a result we have a mostly "desktop" or PC based editing community writing and maintaining Wikipedia for what is now a majority Mobile audience. So calls to replace the default desktop skin with the mobile one are likely to be resisted as they would greatly reduce our recruitment of new editors and put the editing community into decline. Eventually that would either have to be reversed, either by finding a new way to recruit editors or by going back to the compromise of Vector, or even the more editor focussed Monobook as the reader skin. People who only concern themselves with the short term needs of readers may have no doubt as to the best skin to use, but without editors there would be nothing to read. The real challenge is how to make the mobile skin more editor friendly. ϢereSpielChequers 10:08, 27 August 2020 (UTC)
    WereSpielChequers, I'd go a bit further than that. Right now, editing is mostly a text-based activity. Even with the VisualEditor, that's just a nicer way to enter text. Mobile (by which I mean things lacking a dedicated text keyboard) fundamentally sucks at entering large amounts of text.
    The next big advance is editing wikipedia on mobile will not be tweaks like mobile-friendly skins, but a fundamentally new way to interact, such as voice recognition. Voice is making progress, but it's a long way from being commodity software. There are even better technologies in the works, but they're even further away from being ready.
    And, yeah, DataAthlete, for somebody with 3 edits to their name, you might want to tone it down a little. -- RoySmith (talk) 17:38, 27 August 2020 (UTC)
    @DataAthlete: Regarding your comments the paternalistic attitude of older Wikipedia editors that want to force us, readers, to follow their interface preference and I want English Wikipedia editors to stop forcing English Wikipedia readers to use the ugly interface - interface changes are nothing to do with Wikipedia editors (among whom are everybody who has ever posted to this page, including you and me), they are made by the MediaWiki software developers. Their tasklist/noticeboard/discussion forum is at phabricator:. --Redrose64 🌹 (talk) 19:39, 27 August 2020 (UTC)
    That's a bit disingenuous. Many UI changes that at least I would consider as positive have been rejected by the community. I'm sure you can list them off also. The complaint holds some truth in the regard that the community has unfortunate keys to that door. --Izno (talk) 20:31, 27 August 2020 (UTC)
    @RoySmith In general, yes. Mobile and especially smartphones are slower for data entry than keyboards. My experience of the Mobile site is not recent, and it is possible it has been transformed without my noticing. But I think that unlikely as my understanding of Newbies and editor stats is that we are overwhelmingly a stable keyboard using community. I recently spent a while on Quora which is also a very text based site, but I had no problem using that on a tablet. I'm loathe to edit Wikipedia on a tablet, and when I do the first thing i do is switch to the desktop view. I suspect that our best short term route would be to go from two sites to three - smartphone, tablet and keyboard. Longterm I'm sure someone will crack voice recognition, my own first experience with that was over twenty years ago and at the time it seemed a few years away from usable. When and if it ever does become viable you then hit the problem that not everyone has a private office to edit in. It isn't just libraries where voice recognition may not be entirely viable. ϢereSpielChequers 21:52, 27 August 2020 (UTC)
    WereSpielChequers, I have it on good authority that voice recognition will be perfected soon. -- RoySmith (talk) 22:18, 27 August 2020 (UTC)
    Excellent, thanks for settling that beyond all possible response. I look forward to it, and also to the teleport that cometh with it. The prophets who gave us interracial kisses, satellite phones and automatic doors clearly know more of the future than I do. ϢereSpielChequers 18:55, 28 August 2020 (UTC)
  • I translated @DataAthlete's complaint, and it came out as:
I, DataAthlete, prefer prefer the mobile interface, so I use that. I demand that all you useless idiots force everyone else to follow my choice.
For some reason, this hasn't produced a positive response. I wonder why? --BrownHairedGirl (talk) • (contribs) 10:34, 28 August 2020 (UTC)
I love snark that needs no tags. -- llywrch (talk) 22:12, 1 September 2020 (UTC)

javascript being loaded from multiple WMF projects?

I just noticed in my Chrome Developer Tools window, I'm loading a bunch of js files from various WMF projects: 

https://commons.wikimedia.org/w/index.php?title=MediaWiki:Gadget-HotCat.js&action=raw&ctype=text/javascript
https://de.wikipedia.org/w/index.php?title=MediaWiki:Gadget-revisionjumper.js&action=raw&ctype=text/javascript
https://meta.wikimedia.org/w/load.php?modules=ext.gadget.MoreMenu
https://www.mediawiki.org/w/index.php?title=MediaWiki:Gadget-DotsSyntaxHighlighter.js&action=raw&ctype=text/javascript
https://www.mediawiki.org/w/index.php?title=MediaWiki:Gadget-modrollback.js&action=raw&ctype=text/javascript

I've never noticed these before. They don't get loaded in an incognito window, but blanking my common.js doesn't prevent them from being loaded. So, I'm assuming something changed in some gadget I have enabled. Any ideas? -- RoySmith (talk) 12:48, 2 September 2020 (UTC)

I don't think anything changed; these gadgets (HotCat, etc.) are just loaded from external Wikimedia sites. Based on their edit history (for example, HotCat's), they've done so for years now, since 2012 or before. Perhaps Chrome has made a change with how it displays external scripts? Writ Keeper ⚇♔ 13:27, 2 September 2020 (UTC)
RoySmith, some of the en.wp gadgets are proxy gadgets which load code from other wiki's yes. This is to ensure centralised maintainability of these gadgets. If you had any of these gadgets enabled before, then you simply never realised this was happening, it is not new. —TheDJ (talk • contribs) 13:29, 2 September 2020 (UTC)
TheDJ, OK, thanks. It's certainly possible that I've just never noticed it. -- RoySmith (talk) 13:38, 2 September 2020 (UTC)

Reverting edits

Hi, recently a user made vandalism on my article by subverting names of the title. Due to some technical restrictions, I cannot revert his/her destructive edit; while some expert users use Twinkle. How to use this technique? The Supermind (talk) 13:45, 2 September 2020 (UTC)

The Supermind, Give us a hint: which article are you talking about? A link to the page, and specifically which change you're talking about? -- RoySmith (talk) 14:21, 2 September 2020 (UTC)

Kiev Day and Night. Now I fixed manually. User:Bozhena last edition. The Supermind (talk) 14:34, 2 September 2020 (UTC)

watchlist - open or filled bullet at left of list

Despite diligently studying Help:Watchlist, I cannot work out what determines the filled or unfilled bullet at the start of each line in a watchlist. I am figuring it must signify something - but I have no idea what. Can someone help me out?ThoughtIdRetired (talk) 15:38, 2 September 2020 (UTC)

@ThoughtIdRetired: depending on the version of the watchlist you display this may look different - but it should be signifying if the page has changed or not since you last read it - is that what you are experiencing? — xaosflux Talk 15:49, 2 September 2020 (UTC)
Thanks - I should have thought of that. I cannot check if that is what is happening as I have been in and out of my watchlist (and articles on it) trying to work out what was going on. I'll let the dust settle and check for this behaviour in a day or so.ThoughtIdRetired (talk) 15:55, 2 September 2020 (UTC)

Double redirects

I have noticed that double redirects take way too long to fix. At the least, they can take several days. At most, they can take months or longer. This is a problem. We need a faster system to detect and fix them. The current bots are very slow, and the special page is slow to update. Is there any way that we can update the list faster? A bot should be created that looks at any newly created redirects. If they are double redirects, it will fix them instantly. This should be technically possible, after all, SineBot is able to monitor every talk page and sign posts almost instantly. Why doesn't such a bot exist? — Preceding unsigned comment added by I-82-I (talk • contribs) 01:53, 28 August 2020 (UTC)

@I-82-I: WP:BOTREQ is the place to ask someone else to build a bot. — xaosflux Talk 11:47, 28 August 2020 (UTC)
In my experience double redirects are fixed by bots in a matter of minutes, at most hours. EmausBot is currently very active. Can you name specific examples of double redirects which weren't fixed for "several days, ... months or longer"? I find it most likely the problem isn't that the bots are slow but that there are specific issues with such redirects that make them difficult for the bots to detect. Nardog (talk) 03:21, 29 August 2020 (UTC)
I found that double-redirects caused by page moves are fixed up quite fast; typically the bot interrupts me in the middle of me fixing them up manually after a move. However, redirects accidently created pointing to another redirect instead of the actual target page often take days or weeks to be fixed. So, perhaps the bot needs some more hooks triggering its fixup work. --Matthiaspaul (talk) 12:36, 2 September 2020 (UTC)
Yes, double redirects resulting from moves get fixed fast. In fact too fast, and that leads to problems. – Uanfala (talk) 17:57, 2 September 2020 (UTC)

Something is inserting junk into citations

I have seen several edits like this where someone creates a citation that has nonsense information, usually for authors. Other examples have things like "first1=James" with "last1=Cameron,Lewis,Rebekah,Schultz,Peter,New York Times columnist Web exclusive" or somesuch. I have asked the editor who added this why and how he did and he ignored me. Can anyone figure out how this is happening and how we can avoid it? ―Justin (koavf)❤T☮C☺M☯ 20:56, 1 September 2020 (UTC)

"How this is happening" is called Visual Editor. It has over 2,000 open tasks in Phabricator, including many bugs that were reported over three years ago (e.g. T162291 and T174303) and that still cause gnomes to have to clean up after editors who don't know that they are using software that is so buggy. We can avoid it by going to Preferences - Editing and checking the box labeled "Temporarily disable the visual editor while it is in beta". Welcome to the party! – Jonesey95 (talk) 04:02, 2 September 2020 (UTC)
gnomes have always done cleanup work. beginning editors generally don't fully write out references... we like references don't we ? —TheDJ (talk • contribs) 09:13, 2 September 2020 (UTC)
It is not VisualEditor, the same incorrect names are produced for [5] when using the RefToolbar in the wikitext editor. The problem is with the Citoid/Zotero service which attempts to automatically extract metadata from a URL. It probably needs some fixes to the translator for Der Spiegel as described at mw:Citoid#Troubleshooting_the_citoid_service. the wub "?!" 09:55, 2 September 2020 (UTC)
BTW. Anyone can test what the citation extraction platform (backed by Zotero) generates, by using the Wikimedia api directly as mentioned on that page. And Zotero (also used by many researchers around the world to keep track of sources when writing their papers) is always looking for contributors to their translators, to help both Wikipedia and all researches around the world to have better output ;) —TheDJ (talk • contribs) 12:20, 2 September 2020 (UTC)
I wasn't able to figure out the problem with the Der Spiegel translator myself, so have created an issue at the Zotero project [6] for someone to look into it. @Koavf: if you have more examples, please share them and I'm happy to take a look and file further issues if required. the wub "?!" 21:34, 2 September 2020 (UTC)
The wub, I have seen several and they are all of the same sort. If it's really worth tracking down repeats of the same kind of problem, let me know. ―Justin (koavf)❤T☮C☺M☯ 23:29, 2 September 2020 (UTC)

Linter?

The "Page information" menu option sometimes reports Lint errors. So I installed WPCleaner and gave it a try.
The first page I tried was Wikipedia:WPCleaner/Installation. Page information reports 1 lint error. WPCleaner said it was OK. The embedded function in WPCleaner to use CheckWiki also said it was OK.

— What is CheckWiki?

— How do we ask Page information to show us the errors that it has found? — GhostInTheMachine talk to me 20:41, 3 September 2020 (UTC)

The Page Information lint error count is unreliable (see T246403). To see Lint errors on each page, install User:PerfektesChaos/js/lintHint. You will then see a yellow LintHint button at the top right of each page. It will generate a list of errors. If you click it in Preview mode, it will generate a list of errors, along with an arrow that takes you to each error's (approximate) location in the text.
Activating a syntax highlighter, like the one in Preferences - Gadgets, also helps. Feel free to post at Wikipedia talk:Linter if you have further questions; there are a handful of us gnomes working on fixing the errors. – Jonesey95 (talk) 21:09, 3 September 2020 (UTC)
Thanks. LintHint does help. — GhostInTheMachine talk to me 21:39, 3 September 2020 (UTC)

Inputbox directing to editing user's page?

For context, see Wikipedia_talk:Teahouse#Edit_Request_Wizard. Sam-2727 put together an edit request wizard (= a user-friendly form to make edit request).

If a user wants to make a paid editing disclosure, they are brought to Wikipedia:Edit_Request_Wizard/Paid. The page will then bring them to User:(whatever they filled the box with) and preload a {{paid}} template. This means both that they could go to the wrong page if they type the incorrect username (in which case there is no easy clean-up), and also that they have to type something the software probably could take care of. Is there a way to link to the viewing user's page? (Assume a logged-in user if needed.) TigraanClick here to contact me 08:34, 3 September 2020 (UTC)

@Tigraan: I haven't looked at the mechanics of this at all, but Special:MyPage always links to your own userpage. — xaosflux Talk 11:47, 3 September 2020 (UTC)
@Tigraan: OK so that's not going to work for that - a withJS link may work, depending on how people get to that page - if they only get to it from another link a javascript could be invoked to help fill out that form. Would editors ever go to that page directly, or always land somewhere else first? — xaosflux Talk 11:49, 3 September 2020 (UTC)
What's wrong with a button like this: Disclose payment? Nardog (talk) 13:29, 3 September 2020 (UTC)
@Nardog: that should work for a button-based solution (and there is also Special:MyTalk if this needs to target the userpage) - and I'm not seeing they really need a "form" for this so should be the simplest. — xaosflux Talk 13:33, 3 September 2020 (UTC)
Nardog and xaosflux, thanks for the simple solution! I forgot that special page existed. Sam-2727 (talk) 23:17, 3 September 2020 (UTC)

SUL, mediawiki.org and mw:WP:TNT

Tracked in Phabricator
Task T226797

I just visited www.mediawiki.org where I have a handful of contributions from a few years ago. Why does it show me as not logged in? Is it part of WP:SUL and it's just some glitch that I'm not logged in there?

I was looking at mw:WP:TNT which is a redirect to mw:Multilingual Templates and Modules and that leads me to another question: why does diff at Module:No globals show the funky w>DiBabelYurikBot without the normal "(talk | contribs)" links? That's also in the page history. Johnuniq (talk) 03:07, 3 September 2020 (UTC)

@Johnuniq: mediawiki.org is part of SUL, so you should be logged in there. When you first go to a new wiki, you often have to refresh the page before it fully takes effect, though. As for the weird username and missing talk/contribs links, it looks like that's an artifact of that particular revision being imported, rather than actually being made as an edit on this wiki. Jackmcbarn (talk) 04:45, 3 September 2020 (UTC)
Thanks, strange... Johnuniq (talk) 05:17, 3 September 2020 (UTC)
Johnuniq, browsers have started significantly restricting the presence of cookies on other domains. Unless you visit every top level domain in the wikimedia family at least once a day, there is no way any longer that browsers reliably support cross domain authentication like our SUL. You can thank all the ad trackers etc for ruining it for the ‘good’ people. —TheDJ (talk • contribs) 06:57, 3 September 2020 (UTC)
Isn't the internet wonderful, thanks. So if I needed to do something at mediawiki.org, I would have to log in using my normal SUL name and password, and I should expect to have to do that again if returning after a delay of a day or two. Johnuniq (talk) 07:59, 3 September 2020 (UTC)
@Johnuniq: this could be browser-specific, for example I just visited various WMF sites I rarely go to (e.g. ukquote) and was logged in just fine. Are you using "remember me" for your logon on here? — xaosflux Talk 19:11, 3 September 2020 (UTC)
  • This happens to me too. Ugh, I cannot even count how many times I've accidentally edited on Wikidata while logged out. I've found the SUL issue usually happens more often when using the desktop view on a mobile device. Steel1943 (talk) 19:15, 3 September 2020 (UTC)
    @Steel1943: are you using "remember me" on your logon? When this happens, what do you have to do to get logged in (e.g. reload the page, actually enter your credentials, something else)? — xaosflux Talk 19:18, 3 September 2020 (UTC)
    • @Xaosflux: "are you using "remember me" on your logon?" Yep. "When this happens, what do you have to do to get logged in[?]" "Actually enter my credentials". Steel1943 (talk) 20:30, 3 September 2020 (UTC)
      • @Steel1943: do you mind sharing what browser you are using? Can you duplicate this problem with a different browser? — xaosflux Talk 21:59, 3 September 2020 (UTC)
      • @Xaosflux: Arg!!!! I'm using Safari. I've heard about various issues with Wikipedia on Safari, and I guess this takes the cake: I could not replicate the issue on Chrome. On Chrome, SUL worked just fine. Steel1943 (talk) 22:53, 3 September 2020 (UTC)
    If login cookies are misbehaving, try an explicit Special:UserLogout - this is performed server-side, but will invalidate all of your login cookies on all devices, no matter how old or how corrupt they may be. Then you can use Special:UserLogin to create a fresh, new, clean set of cookies. --Redrose64 🌹 (talk) 19:31, 3 September 2020 (UTC)
    • @Redrose64: Just did that, and then went to Wikidata ... which shows me as logged out. I'm just so used to it anymore that it doesn't phase me. Page refreshes don't work either; I have to manually log into each site ... which is weird because oppositely, after I log out of one Wikimedia site, I get logged out of all Wikimedia sites. Steel1943 (talk) 20:32, 3 September 2020 (UTC)
      Special:UserLogout is supposed to log you out of all WP:SUL sites, because it invalidates all of your login cookies. Until a couple of years ago, the login cookie was generated on whichever wiki that you used Special:UserLogin at; nowadays it's login.mediawiki.org that generates the login cookie. If you go to another site - say commons: - and find that you're not logged in, refreshing the same page multiple times is usually nonproductive; but it's often quicker (if not easier) to follow a few links on that site until you hit a page that shows you as logged in. Try the site's "random page" link a few times. --Redrose64 🌹 (talk) 21:57, 3 September 2020 (UTC)
      I figured the logout scenario was intentional. And I shouldn't have to hit "random page" a bazillion times to make my SUL carry over. In fact, I tried that, and after about 10 times, it didn't work. But it did work after time 15. Steel1943 (talk) 22:50, 3 September 2020 (UTC)
Looks like this is a known issue with Safari (phab:T226797). — xaosflux Talk 00:05, 4 September 2020 (UTC)
See also User talk:MusikAnimal/Archive 41#Using steward access?. * Pppery * it has begun... 00:52, 4 September 2020 (UTC)

Milli Vanilli

I am wondering how a vandalized Wikipedia article has ended up in my Wikivoyage userspace with no trace of it being moved there. Several of us at Wikivoyage have investigated and can't find out how this happened. It seems to have happened in August of this year. The only connection that I can see is that a few weeks ago I read the Wikipedia article on my phone, but I didn't edit it. Any thoughts? Ground Zero | t 20:22, 1 September 2020 (UTC)

Ground Zero, Take a look at your post: it's not quite grammatical and the link seems broken. ―Justin (koavf)❤T☮C☺M☯ 20:57, 1 September 2020 (UTC)
Koavf, the links work for me, and I don't see a grammatical error. What is it that you don't understand? Ground Zero | t 21:12, 1 September 2020 (UTC)
Ground Zero, An IP fixed the link and I misread you. ―Justin (koavf)❤T☮C☺M☯ 21:36, 1 September 2020 (UTC)
How bizarre. There you've got the whole history of our Milli Vanilli article up to a certain point in August 2010, almost exactly a decade ago. Whatever miracle caused this, it can't have occurred in 2010 as Wikivoyage didn't exist at the time. – Uanfala (talk) 21:26, 1 September 2020 (UTC)
I did a query on Quarry to find the page id and did another query to find pages with the ids either side of that.
use enwikivoyage_p;
select *
from page
where page_id >= 179513
and page_id <= 179521;
Which returned the following results:
page_id page_namespace page_title
179513 0 North_Vidzeme_Biosphere_Reserve
179514 0 New_York_(NY)
179515 1 New_York_(NY)
179516 1 North_Vidzeme_Biosphere_Reserve
179517 2 Ground_Zero/Milli_Vanilli
179521 3 Traveller1975
voy:North_Vidzeme_Biosphere_Reserve is in the import log as being imported on 13 July 2020 and voy:User talk:Traveller1975 was created on 14 July 2020, so whatever it was, it happened sometime between those two items. -- WOSlinker (talk) 22:58, 1 September 2020 (UTC)
As an importer here myself, I know that the import tool can be weird when trying to work with many revisions at a time ... it probably timed out and never logged properly. Ground Zero, is there some sort of gadget/tool (on Wikipedia/Wikivoyage's side)that facilitates easy imports to Wikivoyage that you might have enabled somewhere and perhaps misclicked on? I can't find any evidence of that but that's the only plausible explanation I can think of. Re timestamps: revision ID numbers can tell an interesting story too ... the first revid after your imported edits is 21:13, 13 July 2020 (UTC) and the last one from before your first imported edit is from 21:08 (UTC) on the same day. Graham87 04:09, 2 September 2020 (UTC)
Wasn't voy: started off as a kind of mirror/fork of WikiTravel? Maybe the route was Wikipedia → Wikitravel → Wikivoyage. --Redrose64 🌹 (talk) 18:32, 2 September 2020 (UTC)
@Ground Zero: The revision ID of the latest imported revision of voy:User:Ground_Zero/Milli_Vanilli is 4020727. Revision 4020728 is a talk page comment, from you, about a failed attempt to import a page into your userspace. So it looks like the import aborted mid-way through, at the vandalized revision, and didn't log properly, but it really was you. Suffusion of Yellow (talk) 22:03, 2 September 2020 (UTC)
still funny that it hit this particular article. ---Sluzzelin talk 22:13, 2 September 2020 (UTC)

@Uanfala, WOSlinker, Graham87, Redrose64, Suffusion of Yellow, and Sluzzelin: This does seem to have happened around the time that I was experimenting with the Import tool in order to deal with Wikipedia articles in the Category:Copy_to_Wikivoyage. I'm not a very technical person, and had forgotten that I had been using that, so it makes sense that I may have started the import by accident, then cancelled it part way through. Thank you for taking the time to investigate or at least think about this. Ground Zero | t 01:29, 4 September 2020 (UTC)

Edit preview not working correctly?

Anyone experiencing issues with edit previews not working correctly? I was making these changes, converting stuff like "Runner Up" to "runner-up". When I'd click Preview, the preview window showed that they were still capped and unhyphenated, even though the edit window showed that they were lower-case and hyphenated. the Show changes feature indicated that the change had been made. I also changed "Evicted By Housemates", to "Evicted by housemates", and after hitting preview, only one of the instances changed to lowercase. The others were still capped. So, just curious if anyone else is experiencing this. I'll blow out my cache and whatnot. Cyphoidbomb (talk) 16:05, 3 September 2020 (UTC)

1. I haven't had such an experience like that (lately/today/at all)
2. Are you sure you were looking at the right "Runner Up"s? I see 7 instances on that page, where you're changing only 5 of them
3. YIKES!! What devices do readers of this article use? One tables at 80%, the next is at 60%, and the broken <small> tagging for the Notes makes the entire Reflist teeny. Please, oh please take a look at WP:ACCESS, and rework the tables (and everything else) to not go below 85%. — JohnFromPinckney (talk) 19:48, 3 September 2020 (UTC)
@JohnFromPinckney: Hey, I always leave a possibility open that I'm an idiot and I'm missing something.   I'm looking at this version of the article, (current as of this moment,) and in the Housemates Status section (miscapitalised), which is the section I edited, I still see "1st Runner-Up" and "Walked (5th Runner-Up)" and "Health Issue", even though I fixed those in my edit. I just purged that article's page cache and my browser cache, but still see the caps. I've got to be missing something here. It's also not something I've ever seen in page preview mode. Also, good luck making sense of the Bigg Boss templating across all the Indian language versions... Hindi, Tamil, Malayalam... Cyphoidbomb (talk) 00:22, 4 September 2020 (UTC)
OK, having looked closer at this, I think Template:Runner-up and Template:Quit are artificially forcing capitalisation when it's not warranted or needed. Cyphoidbomb (talk) 01:50, 4 September 2020 (UTC)
@Alucard 16: Hey there, since I see you have some knowledge of these templates, any thoughts on this? Since most of the stuff we add to tables seems to be more "Sentence case" formatted, I'm not exactly sure that "Runner-Up" is quite right from that perspective, although I'm not arguing with any Big Brother/Bigg Boss consensus(es) in mind. But if someone provides an explanation for a character leaving, like "Health issue" that probably doesn't need to be capped, right? Not trying to nitpick, but in a table like wot you see at Bigg_Boss_(Hindi_season_13)#Housemates_status, there's somewhat of a inconsistency there that I'm not sure how to encourage an entire community of casual editing series fans to adhere to. Thanks and regards, Cyphoidbomb (talk) 02:13, 4 September 2020 (UTC)

Rangeblocked

Hi I'm signed in right now but on my cell phone while not logged in I was looking at Wikipedia and looked at the edit page and noticed it said I was rangeblocked until Feb. 2021. I have only edited one thing ever on Wikipedia, it was when I saw a typo,the word because misspelled and so I excitedly fixed it. I searched a long time for minor typo that I could edit and become a part of Wikipedia. But I don't know if that would cause a rangeblock. I'm wondering what could've happened? — Preceding unsigned comment added by CircleRightEye (talk • contribs) 10:17, 4 September 2020 (UTC)

@CircleRightEye: See Template:Rangeblocked. --Malyacko (talk) 11:24, 4 September 2020 (UTC)
@CircleRightEye: Most cell phone IP addresses are blocked from editing Wikipedia. I personally think that it's overzealous, but I've lost that argument here before. --Ahecht (TALK
PAGE) 13:05, 4 September 2020 (UTC)

Template to fast-click to bottom of a thread?

Do we by any chance have a template for use in the edit window, whereby an editor can quickly go to the bottom of a a section? Seems to me that these long drawn-out discussion topics could use such a tool. What a pain to scroll and scroll to get to the bottom to post a comment. — Maile (talk) 13:38, 4 September 2020 (UTC)

Ctrl+End? -- Michael Bednarek (talk) 14:02, 4 September 2020 (UTC)
Thank you! — Maile (talk) 14:31, 4 September 2020 (UTC)

question about gadget definition

This might be a stupid question but i need to make sure. On bnwiki we have a UTCLiveClock gadget (enwiki also has it) & a LocalLiveClock gadget (same code with little variant). Recently i updated their code and definition (by copying from enwiki). Here is the diff, notice that i used UTCLiveClock-pagestyles & UTCLiveClock.css for both gadgets. Is it fine? will it cause any problem? are css code going to be loaded twice? --আফতাবুজ্জামান (talk) 01:55, 4 September 2020 (UTC)

If the same CSS file is loaded twice, that's generally harmless. If something happened in between the two loads to modify the effects of the first load, those modifications might be reversed by the second load. --Redrose64 🌹 (talk) 08:03, 4 September 2020 (UTC)
@Redrose64:Thanks for reply. This two gadgets aren't default. Unless some enable both at the same time, i'm guessing there wont be any problem. Right? --আফতাবুজ্জামান (talk) 14:36, 4 September 2020 (UTC)

Parsing wikitext in javascript?

What's the best way to parse wikitext in javascript? I'm thinking something with the functionality of mwparserfromhell. What I'm doing now is just regex matching to find the templates I need, but that broke when somebody formatted a template in a slightly different (but still legal) way. This, of course, is unsurprising, since everybody knows that trying to abuse regex as a language parser is a lost cause. -- RoySmith (talk) 13:32, 3 September 2020 (UTC)

@RoySmith: See mw:Alternative parsers. I've used wtf_wikipedia before, and the API is pretty simple. Vahurzpu (talk) 16:21, 3 September 2020 (UTC)
Vahurzpu, Cool, thanks. I'll give wtf_wikipedia a shot, if only for the awesome name. -- RoySmith (talk) 17:09, 3 September 2020 (UTC)
@RoySmith: Do you actually need to do the parsing in JavaScript, or do you just need JavaScript to be able to get a parsed result? If the latter, consider using action=expandtemplates or action=parse in the MediaWiki API, or /transform/wikitext/to/html in the Parsoid API (this is how I designed WP:EPH, and I did it that way specifically because trying to parse myself led to the exact problem you're having now). Jackmcbarn (talk) 20:22, 3 September 2020 (UTC)
Jackmcbarn, I need to get the templates, and the template arguments. As far as I can see, action=parse will give me a list of templates, but just that a template exists, i.e.:
"templates": [
    {
        "ns": 10,
        "exists": "",
        "*": "Template:Sockpuppet"
    },
but not what the template arguments are. Am I just missing something? It seems obvious that there would be a way to get the template arguments, but I'm not seeing it. -- RoySmith (talk) 21:02, 3 September 2020 (UTC)
You can use prop=parsetree, and then deal with it in xml instead? —Cryptic 21:58, 3 September 2020 (UTC)
@RoySmith: If you want arguments, you almost certainly want Parsoid. That's exactly what EPH does; have a look at it. Jackmcbarn (talk) 01:31, 4 September 2020 (UTC)
Jackmcbarn, Yeah, the more I look at this, the more I'm convinced you're right, but man, what a strange way of doing things. The arguments in
{{sockmaster|blocked}}
and
{{sockmaster|1=blocked}}
parse totally differently ([7] vs [8]). -- RoySmith (talk) 01:53, 4 September 2020 (UTC)
@RoySmith: That's still MediaWiki's parser. Parsoid gives you much more sane parameters: {"1":{"wt":"blocked"}} for Jimchu23 and {"1":{"wt":"Viralvilla"},"2":{"wt":"proven <!-- blank/blocked/proven/confirmed -->"},"evidence":{"wt":"<!-- wiki link to alternate location of evidence if it's not on the master's SPI page -->"},"locked":{"wt":"no"},"notblocked":{"wt":"no"},"altmaster":{"wt":""},"altmaster-status":{"wt":"<!-- blank/blocked/proven/suspected -->"},"length":{"wt":""},"cat":{"wt":"<!-- Custom category name for the sockpuppet -->"}} for Ultrabone. Jackmcbarn (talk) 01:57, 4 September 2020 (UTC)
Jackmcbarn, What URL did you use for the above? I tried https://en.wikipedia.org/api/rest_v1/page/pagebundle/User:Ultrabone, which just gets me a not_found error. -- RoySmith (talk) 02:58, 4 September 2020 (UTC)
@RoySmith: I used https://en.wikipedia.org/api/rest_v1/page/html/User:Ultrabone. Jackmcbarn (talk) 03:02, 4 September 2020 (UTC)
Jackmcbarn, That gets the HTML. How did you get the JSON you showed above?
The documentation is baffling. this talks about routes with /v3 in them. this has a totally different set of routes. I'm assuming the later is machine-generated from the server's actual routing tables, so that's probably the correct one. Some of those routes (for example, the /html/ one you give above) work, but the /data-parsoid/ one doesn't. -- RoySmith (talk) 03:11, 4 September 2020 (UTC)
@RoySmith: That JSON was in the HTML, specifically in the data-mw attribute of the outer element. Jackmcbarn (talk) 03:59, 4 September 2020 (UTC)
Jackmcbarn, Wow. That's, um, a non-intuitive way to design this API. -- RoySmith (talk) 15:45, 4 September 2020 (UTC)

unwanted transclusions

Hello,


I am wondering how to get rid of unwanted transclusions, for instance the page Relics of Sariputta and Moggallana has an transclusion in the first part of the Sanchi relics section, which makes no sense being there. I can't find anything in the code that would cause this so i am curious how to get rid of it. Wikiman5676 (talk) 06:46, 4 September 2020 (UTC)

@Wikiman5676: I'm not seeing any transclusion in that section. What specifically is transcluded? --Ahecht (TALK
PAGE) 13:15, 4 September 2020 (UTC)
@Ahecht: it's the subsection the says sanchi relics under the excavation of lost relics section. From "In 1851..." to the quote by cunningham that begins "Sariputra and Mahamoggallana...". If you try to select it via visual editor it will lump the texts together as a transclusion. Wikiman5676 (talk) 15:15, 4 September 2020 (UTC)
I don't use VE, but I can replicate this problem. Edit the section using this link, and it highlights the first paragraph of the "Sanchi relics" section along with the indented quote, using a gray background. There is a pop-up box overlaying the gray-highlighted area, with text that says: "Transclusion [new line] Generated from: Sfn, Sfn, Sfn, Sfn, Sfn, Sfn, Quote". The pop-up has an Edit button. When you click the Edit button, another pop-up window appears, allowing you to edit (some of?) the contents of the gray-highlighted area. It is bizarre. I don't see any errors on the page that would lead a parser to misinterpret this section. – Jonesey95 (talk) 15:53, 4 September 2020 (UTC)
@Jonesey95 and Wikiman5676: Inserting a line break before the {{quote}} seems to have fixed it (see Special:Diff/976720114). --Ahecht (TALK
PAGE) 16:05, 4 September 2020 (UTC)
Thanks! Wikiman5676 (talk) 16:59, 4 September 2020 (UTC)

Query about the look and feel of fr-wiki

Preview fr.wp in Vector skin

Not sure where the best place to ask this is, but I went to French Wikipedia today and found that it looks radically different from how Wikis usually look: [9] [10]. Initially I thought I'd landed on the mobile version by mistake! I don't think I like that look at all, the text appears too narrow and I hope we never go down that route here at enwiki... Anyone know what the rationale for the change was? Cheers  — Amakuru (talk) 11:55, 21 August 2020 (UTC)

Could you post a screenshot of what you're seeing (maybe side-by-side your en view)? I see the font a bit smaller but nothing too radical. –xenotalk 11:57, 21 August 2020 (UTC)
Wikipedia:Village pump (technical)/Archive 183#New look? Nardog (talk) 12:16, 21 August 2020 (UTC)
It looks very different to me too. The most striking difference is a lack of borders separating the article/othercontent from the sidebar, and the replacement much larger whitespace. CMD (talk) 12:26, 21 August 2020 (UTC)
The difference is best seen on a wide screen. Currently English Wikipedia fills the full width, while the French Wikipedia seems to have a fixed maximum width. It does seem to be skin dependent and associated with the planned Vector upgrade. The preferences section has a note that says legacy Vector will still be available as an option. —  Jts1882 | talk  12:39, 21 August 2020 (UTC)
Ah, that's why I don't see a change. –xenotalk 12:41, 21 August 2020 (UTC)
The French Wikipedia is currently testing out the new desktop improvements. Sam Walton (talk) 12:44, 21 August 2020 (UTC)
(edit conflict) This is mw:Reading/Web/Desktop Improvements, discussed a bit back at WP:Village pump (technical)/Archive 183#New look?. LittlePuppers (talk) 12:47, 21 August 2020 (UTC)
  • Thanks all for the replies. So it sounds like this is going to be foisted on us at some point next year, whether we vote for it or not... Has anyone confirmed that the updated format really is beneficial? It's easy to say we should just stick with the old format, but honestly I don't see the benefit of restricting the width of the readable text. Just means more scrolling, and the layout of the other links looks more messy too. Comparing to other "modern" websites is misleading, because on most websites there isn't the same volume of text that you find here, and comparing to the mobile view is a very poor idea as the mobile view is already crap. I don't even use it on my mobile. 😏  — Amakuru (talk) 18:46, 21 August 2020 (UTC)
    Don't worry too much, there have already been several suggestions for a user script to fix the limited width. The smaller logo is good, but the option to hide the sidebar menu is just plain silly. I hope that there will be further changes yet... — GhostInTheMachine talk to me 18:58, 21 August 2020 (UTC)
    You don't need a script, a few CSS rules are sufficient. I provided one at Wikipedia:Village pump (technical)/Archive 183#New look?, but it's not enough on its own. You need all of these: 
    .skin-vector-max-width .mw-page-container {
  min-width: none;
  max-width: none;
  padding: 0;
}
.skin-vector-max-width .mw-workspace-container {
  max-width: none;
}
.skin-vector-max-width .mw-content-container {
  max-width: none;
}
    As mentioned before, to alter the widths for French Wikipedia only, these rules go in fr:Special:MyPage/vector.css (if you have one) or in fr:Special:MyPage/common.css (if you don't); or to alter for all WMF sites, put them in m:Special:MyPage/global.css. --Redrose64 🌹 (talk) 08:48, 22 August 2020 (UTC)
  • Has anybody realised that the reason that many websites have a broad white margin is so that there is space available for advertising and clickbait? --Redrose64 🌹 (talk) 22:36, 21 August 2020 (UTC)
    There are optimal lengths of lines of text, see Line length. Thanks. Mike Peel (talk) 22:48, 21 August 2020 (UTC)
    The optimal length varies for different people. We shouldn't be forced to use somebody else's optimal length, and if people can't handle long lines, they have the window resize feature available (the technology has existed since the 1980s). Some people who have widescreen monitors may have chosen one in preference to a 4:3 monitor because they wanted to use the extra width - they don't want it wasting with big white rectangles. --Redrose64 🌹 (talk) 08:43, 22 August 2020 (UTC)
    This. I wouldn't object quite so much if there were a way to easily revert to full width for the typical reader who is not logged into an account. 24.151.56.107 (talk) 15:19, 24 August 2020 (UTC)
 
Logo overlaps article talk tabs

The examples provided at mw:Reading/Web/Desktop Improvements show the upper left Wikipedia logo 'overlapping' the Article and Talk tabs. However, the actual implementation does not do this, creating the large whitespace on the left of the article. To make matters worse, the horizontal length of the logo is extended by the positioning of the collapse sidebar button directly to the logo's left. This feels like a really easy thing to pick up on, so was it 1) missed and/or thought irrelevant, or 2) changed or accepted without updating the mediawiki page which should be explaining these changes? Either way this feels like poor form. CMD (talk) 07:06, 25 August 2020 (UTC)

Please whoever have a decision power here at Wikipedia don't let that design change happen. It is terrible, the limited width is very bad on large monitors. Making things very hard to read even with glasses. I've been searching for a couple of weeks now how I could remove it. I first though it was a bug and I landed on the mobile version. Then has it wasn't going away I though Wikipedia FR somehow got hacked. But no it seems it something on your part. At the moment I started to read Wikipedia in other languages to not suffer that design change. But if unfortunately it came to become a general change I don't think I would use Wikipedia as much as before as it would have become quite hard to read for me. - Sincerely a simple and concerned Wikipedia reader. — Preceding unsigned comment added by 87.231.14.108 (talk) 20:34, 30 August 2020 (UTC)

"the limited width is very bad on large monitors. Making things very hard to read even with glasses" That doesn't really make sense btw.. The size of the text is the same, so your ability to read with or without glasses, shouldn't depend on a constraint to width. The limited width should only affect those who want lots of very dense information. I advise anyone with difficulty reading to use the browsers zoom function btw.. It is awesome. ctrl + and ctrl - (or command + and command - if you are on macOS). —TheDJ (talk • contribs) 09:12, 31 August 2020 (UTC)
The reason it is bad is because of increased vertical scrolling, nothing to do with font size. — xaosflux Talk 18:58, 4 September 2020 (UTC)

Need function that returns value of line which it is placed

Is there a magic word or some sort of function that returns the value of the line which it is placed? For example, if it were placed on the top line of an article, it would return the value "1". Steel1943 (talk) 16:49, 4 September 2020 (UTC)

@Steel1943: hmm, interesting. What would you be considering a "line" here though? (Anything before a "line break" is generated? - would a section header count?) — xaosflux Talk 17:15, 4 September 2020 (UTC)
@Xaosflux: Like the line number that shows up when looking at edit diffs or when editing modules. Steel1943 (talk) 17:17, 4 September 2020 (UTC)
To answer the question directly, no. It might be possible to do it, but it doesn't seem like something generally necessary that couldn't be done more directly using the same method. --Izno (talk) 17:32, 4 September 2020 (UTC)
  • So, the reason I asked is I'm trying to write something that determines if it is on the top line of the page ... which is why I was hoping there would be a way to just grab the line number since that would obviously be "1" in this case. Steel1943 (talk) 17:47, 4 September 2020 (UTC)
Lua can do this by reading the unparsed wiki text and looking for the thing you placed there. So if you have a template, {{line marker}}, a lua module called by {{line marker}} can fetch text line-at-a-time and search each line for its parent template as a text string ({{line marker}}). cs1|2 uses something similar when it figures out how to format dates according to {{use xxx dates}} though it doesn't count lines. {{line marker}} then renders the line number returned from the lua module
—Trappist the monk (talk) 18:01, 4 September 2020 (UTC)
@Trappist the monk: That doesn't sound like it does what's being asked for. In particular, if you used {{line marker}} in two different places in a page, it wouldn't be able to tell which one it was being called from. Jackmcbarn (talk) 18:22, 4 September 2020 (UTC)
It was the simple example. If you want the third marker then write the template call with a key: {{line marker|3}} (key could be any plain text). When the module is invoked by that form of the template, it will look for a string that matches "{{line marker|3}}" (perhaps with allowance for whitespace ...). {{line marker}} isn't just an empty template.
—Trappist the monk (talk) 18:39, 4 September 2020 (UTC)
@Trappist the monk: Well okay, but now you have to use a unique key every time, and there's no way to generate one of those either. Jackmcbarn (talk) 18:50, 4 September 2020 (UTC)
Yeah, you do have to manually make sure that the keys are unique though I suppose the module could scan for duplicates as a modicum of error avoidance. Examples hard-coded to look for {{tlx|line marker}}:
{{#invoke:Sandbox/trappist the monk/what line|main}} → Found at line: 717 – as I write this in a section edit, this one reports line 8
{{#invoke:Sandbox/trappist the monk/what line|main|3}} → Found at line: 720 – this one reports line 11
Notepad++ agrees with the line numbers.
Presumably this #invoke could go inside another template and report the line number of that template (I haven't tried this so I don't know what hoops need jumping to make that work – I suspect that a mixture of positional and named parameters as keys would render this scheme unworkable because named parameters can be placed anywhere in the wikitext of a template).
—Trappist the monk (talk) 20:03, 4 September 2020 (UTC)
  • A module could be derived from Module:Is infobox in lead. It scans the wikitext source to see if the named infobox is before the first heading. What sort of issue are you trying to test for? — GhostInTheMachine talk to me 18:10, 4 September 2020 (UTC)
    • I'm specifically trying to run a test to see if something is not on "Line 1", so I don't think any variation of that module would work since the module seems to check if it is placed before the first level 2 section header on a page, and I'm trying to get the test more precise than that. Steel1943 (talk) 18:41, 4 September 2020 (UTC)
      • Maybe the SD template might use a Lua module to see if the {{short description}} really is the first thing in the source. i.e. Test for something in front of it rather than thinking about line numbers. As an alternative, it might be "cheaper" to scan the sources held in a database dump to see if the short descriptions are at the top of their articles. — GhostInTheMachine talk to me 19:03, 4 September 2020 (UTC)
      • Strictly speaking, the {{short description}} does not need to be on line number 1 of the wikitext - it just needs to precede all other templates (in case one of those templates itself tries to set the SD). That seems a simple test to me and I do want an excuse to write some Lua.   — GhostInTheMachine talk to me 19:17, 4 September 2020 (UTC)
        @GhostInTheMachine: Someone's been looking at my recent edits ...   Steel1943 (talk) 19:47, 4 September 2020 (UTC)
@Steel1943: No, and one won't be added. The issue is that the Parsoid team wants to be able to cache pieces of parsed pages, so anything that detects where something is in the page, or what's before or after it, is out. And in fact, when people come up with clever ways to do that kind of thing anyway, the developers consider them bugs and remove them. Jackmcbarn (talk) 18:20, 4 September 2020 (UTC)
Jackmcbarn, this :D —TheDJ (talk • contribs) 18:24, 4 September 2020 (UTC)

Username query with wildcards?

Special:ListUsers doesn't seem to support wildcards. I want to show a list of users with "foobar" in them (i.e., regex .*foobar.*). Can I just query the user table somewhere, or is there a special page or tool that supports wildcards? Thanks. —[AlanM1 (talk)]— 03:51, 5 September 2020 (UTC)

@AlanM1: You can use Quarry. Go to https://quarry.wmflabs.org/ and run select user_name from user where user_name like '%foobar%';. Warning: it will be slow (it took over a minute to return when I tried it). Jackmcbarn (talk) 03:55, 5 September 2020 (UTC)
@Jackmcbarn: Thanks! I did select user_name from user where user_name like '%chandandolui%'; and got two results: Iamchandandolui and Mr.chandandolui. I then did select user_name from user where UPPER(user_name) like '%CHANDANDOLUI%'; which ran for about a minute and returned no rows. I'm a little rusty but what am I missing? Shouldn't I get at least the same two rows? —[AlanM1 (talk)]— 04:18, 5 September 2020 (UTC)
Update: Asking at mw:Talk:Quarry. —[AlanM1 (talk)]— 07:53, 5 September 2020 (UTC)
(Or you could ask at WP:QUARRY here.) Most (all?) text is stored with case-sensitive collation, so LIKE is also case-sensitive. You can either use an explicit collation, like SELECT * FROM user WHERE user_name LIKE '%chandandolui%' COLLATE utf8_general_ci;, or use a case-insensitive regex search, like SELECT * FROM user WHERE user_name RLIKE '(?i)chandandolui';. Both show Chandandolui, Iamchandandolui, Mr.chandandolui, and Chandandolui1; the regex search is more flexible and about twice as fast.
Your attempt with UPPER() failed because user_name is a varbinary(255); see https://mariadb.com/kb/en/upper/ and DESCRIBE user. —Cryptic 10:02, 5 September 2020 (UTC)
@Cryptic: Thanks! The varbinary type did not occur to me. RLIKE is new to me, and very welcome. —[AlanM1 (talk)]— 11:20, 5 September 2020 (UTC)

Pages using infobox song with unknown parameters

I've looked all over I Hope You're Happy Now (Carly Pearce and Lee Brice song), and I can't figure out what's causing it to show up in Category: Pages using infobox song with unknown parameters. What unknown parameter has been inserted that I'm missing? Ten Pound Hammer • (What did I screw up now?) 04:16, 4 September 2020 (UTC)

@TenPoundHammer: It's format. You can see it by editing the page and previewing it. Jackmcbarn (talk) 04:22, 4 September 2020 (UTC)
In preview mode I see: "Warning: Page using Template:Infobox song with unknown parameter "format" (this message is shown only in preview)." At Template:Infobox song a search for "format" brings me to: "The format parameter has been removed as a result of the July–August 2020 talk page discussion. Any information entered in |format= will not be displayed." — JohnFromPinckney (talk) 04:27, 4 September 2020 (UTC)
Wow, that might be cause for a bot run. It filled the unknown parameters cat with nearly 50,000 titles. Raymie (t • c) 23:45, 4 September 2020 (UTC)
Yowsa. Maybe this is a chance for me to learn how (and where) to request bot operations. — JohnFromPinckney (talk) 01:07, 5 September 2020 (UTC)
@JohnFromPinckney: I'll send you to User talk:Primefac to have User:PrimeBOT do this. If it's implementing a discussion outcome, despite the very large article set, I doubt this would be seen as controversial. Raymie (t • c) 03:06, 5 September 2020 (UTC)
Interested editors may want to look at the template's talk page first. The discussion was attended by a small number of editors, and it is unclear to me (the person who eventually removed the parameter based on an edit request) whether the discussion was advertised. It would be unfortunate to have to reverse 50,000 bot edits if significant objections arise only after editors notice the parameter's removal. – Jonesey95 (talk) 13:40, 5 September 2020 (UTC)

Page notices in VisualEditor

The notice appears below the protection status. It's generally collapsed, barely visible, and I've only recently realised they're there at all–I usually just press X straight away, lots of words, and the first few about semi-protection are irrelevant, so who'd bother scroll down to read the rest? Can we get protection info either (ideally) (a) removed from that notice, or (b) placed below the page notice? For BLP articles especially, we should want the notice about BLP highlighted, not "this article has been configured with a mighty semi-protection setting by X for Y until Z on the time A [thank me!]". Or can we get it moved above, like the normal source editor? Can any of these be done locally, or do we need a Phab ticket for any of them? ProcrastinatingReader (talk) 11:24, 30 August 2020 (UTC)

@ProcrastinatingReader: can you provide an example page with this problem (an article would be best)? — xaosflux Talk 12:26, 30 August 2020 (UTC)
Xaosflux, a recent one I came across was Kenosha protests where there's constantly BLP violations, so making sure the notice is visible helps. It's not the most obvious since it only has one notice, but still an issue. On the larger Mac I'm using now I can see the notice in full, but on my 13" Mac I can barely even see the "Notice about sources" part until it cuts off, so how visible the issue is to you might depend on the device you're using. It's most visible, regardless of monitor, on articles like Donald Trump (due to long protection reason & multiple page notices). Where screen real estate is limited we should be prioritising the useful notices first, if you can kinda follow what I'm trying to get at? ProcrastinatingReader (talk) 12:43, 30 August 2020 (UTC)
@ProcrastinatingReader: so you'd want to move the protection message possibly to the bottom - if it is a protected page that you can edit --- however, I'm assuming this is still the most important notice if it is on a page you can't edit - since it tells you why you can't edit it; correct? — xaosflux Talk 13:46, 30 August 2020 (UTC)
Xaosflux, well, I don't think I can open up VisualEditor at all on a page I can't edit, so I don't think any exemption needs to be made for that case. Example: I found Kimberly Klacik, and indeed I can't open up VE at all on here. I can only see a "View source" button and the suggestion to submit an edit request (since you're (I)admin, I think you can only test this in incognito).
For pages I can edit, hence can open VE, what you say is mostly correct & is what I'm getting at. Really, I don't think the protection message should be shown at all on there, except maybe for full protection: the semi/ECP notices aren't helpful there, and anyway we have the coloured lock in the top right before one can click edit. But, if it really must be, then should be at the bottom of the notice list imo. If kept, the message should at least be cleaned up to look presentable. I think the eye naturally skims over that dumped log, especially in a view like VE, at least my eye does. ProcrastinatingReader (talk) 14:12, 30 August 2020 (UTC)
  • I agree that the protection notices are meaningless here. Looking at Kenosha protests, as noted, you can't even get into the editor on a page you can't edit, so the only people who are ever going to see the semi-protected notice are people it doesn't apply to. And the stuff about the protection settings being moved are just noise; I can't think of any way having that information will help somebody edit the page. So, yeah, those two items shouldn't be there. -- RoySmith (talk) 14:20, 30 August 2020 (UTC)
    @RoySmith and ProcrastinatingReader: sure you can, try opening this link without being logged in (use a private browsing/incognito mode tab: link). (It is where you end up from viewing source with the visualeditor if you don't have edit access). — xaosflux Talk 14:28, 30 August 2020 (UTC)
    Xaosflux, sorry, I'm being a bit slow & blind I think, but where's the link to open VE? Whilst your direct link opens up, when I click "view source" I see this. I can't find a link to VE here? And even on your direct link I can't actually edit any words. ProcrastinatingReader (talk) 14:34, 30 August 2020 (UTC)
    At a closer look, the notices on the direct link look screwed. It repeats all the page notices twice (?), and it doesn't show the log entry but instead the "please make an edit request" notice. The direct link just looks bugged, tbh. ProcrastinatingReader (talk) 14:44, 30 August 2020 (UTC)
    Xaosflux, Well, OK, somebody could hand-craft a URL, but the most obvious way to get there is to click in the "Edit" link at the top of the article, but that turns into "Read" if you don't have permission to edit. If you wanted to cover this edge case, then the protection notices should only be shown to people who don't have write permission. -- RoySmith (talk) 14:41, 30 August 2020 (UTC)
  • OK, lets break this down a bit :D Here is a rather clean page to use as an example: testwiki:VE-alert-testing1 (Direct link in to editing mode in debug mode: link). If I'm hearing you right, basically you'd like to move the order of precendence, in the visual editor page alerts, to place protection "warnings" such as (semiprotectedpagewarning) after the other edit notices. This is not managed locally, so yes would require a phab ticket and would be a global mediawiki change (or you could request that there is a per-project customization capability built). I don't think this is going to get general support though - the argument will need to be that a "page notice" is more important in general then a protection notice. Even here on enwiki, the proection notice is consider very important - for example it lets admins (who can edit through protection) know that they shouldn't be editing the page without a good reason. — xaosflux Talk 14:57, 30 August 2020 (UTC)
    Xaosflux, Re, it lets admins (who can edit through protection) know that they shouldn't be editing the page without a good reason, that's something I hadn't considered. Thanks for pointing it out. -- RoySmith (talk) 15:01, 30 August 2020 (UTC)
    Xaosflux, for example it lets admins (who can edit through protection) know that they shouldn't be editing the page without a good reason I agree for that niche case, i.e. pages full-protected due to dispute, and an admin might want to be aware of that before editing, & did hint at it above. Admins editing through full-protection is a very small % of edits, though, but certainly we could keep it shown for full-prot cases only. Even in those cases, there's still the golden lock in the top-right before clicking edit. With your test wiki example, for my eyes, the notice is dwarfed by the log message even when it isn't collapsed due to screen size. It just feels too routine, and hence doesn't demand attention. I think many editors are used to just clicking 'X' habitually. Some kind of per-project customisation capability would be nice. Even if the log message is to be kept, the ability to make it look... less convoluted... would be nice. ProcrastinatingReader (talk) 15:26, 30 August 2020 (UTC)
    @ProcrastinatingReader: hmmm - a possible option without doing too much change (still require a phab ticket) may be to let that logline (the entire ul element for it actually) be collpable and collapsed by default maybe. Do you think that would help? — xaosflux Talk 15:38, 30 August 2020 (UTC)
    Xaosflux, as in with like a [show] to the side? I think that's definitely a big improvement. As a side thought, given how well-designed and low-verbosity VE is, I do wonder how Design approved this in the first place. I do still think even the leftover sentence is too routine and verbose, though, and (since I doubt we'll be able to get the manpower to have them revisit at this issue twice) I'd like to find an easy-to-implement, low controversy, and customisable way to make this work fully off the bat. Ideally some way to either pop it into the config, or let local IAdmins customise it, would be neat. I'll take a flick at the code and ponder on the idea... Should seem doable since it's only the first .ve-ui-mwNoticesPopupTool-item (wholly containing the log entry) I'm trying to mess with, so the option of completely disabling per-project via config seems trivial, but then we'd need to work out something for full-protected-pages... ProcrastinatingReader (talk) 15:47, 30 August 2020 (UTC)
    Xaosflux, I bring good news. VE injects the editnotices in "includes/ApiVisualEditor.php", line 318. There's actually some good design here, because extended-confirmed is treat as "semi-protection" (see here), so it's easy to hide the semi and EC ones and still keep the full protections (and TE, and whatever else) for admins. So, the easy way with, I believe, no collateral is just to add a config param that doesn't add "semi-protected" editnotices. Few lines of code, might even make it through the Phab process sometime this year. The second idea is just to remove the first <div> using JavaScript locally, if it has a "semiprotectedpagewarning" div below it. Has to be JS (can't use CSS) because it's applied to the child object, and we'd have to remove the parent of that div. Both would do the trick. I'd prefer the latter because we can do it sooner, with the caveat that it requires JS, but so does VE so I don't think it's too big of an issue. Note, regarding the non-logged-in example you gave above, neither option will affect the niche case of crafting a URL, per line 277, users who can't edit the page (but see VE) are shown a separate notice, so this has no effect on them. Thoughts? ProcrastinatingReader (talk) 17:50, 30 August 2020 (UTC)
    @ProcrastinatingReader: if those have class wrappers, we could just display:none them locally right? — xaosflux Talk 19:14, 30 August 2020 (UTC)
    Xaosflux, kinda but, issue here is that non-first-child notices all have a border-top associated with them. If we just display:none it, it leaves the border-top from the page editnotice visible. For an idea, edit here and for the first child of .ve-ui-mwNoticesPopupTool-items set display: none, the line stays. Now if you remove the element completely, line goes. Pretty small detail, but... ProcrastinatingReader (talk) 19:28, 30 August 2020 (UTC)
    fwiw, the closest I've gotten with CSS is: 
    #semiprotectedpagewarning, #semiprotectedpagewarning + ul {
  display: none;
}
    And the JS solution being:
    mw.hook( 'wikipage.content' ).add( function () {
	$('#semiprotectedpagewarning').closest('.ve-ui-mwNoticesPopupTool-item').remove();
});
    ProcrastinatingReader (talk) 19:59, 30 August 2020 (UTC)
    @Xaosflux: following up, thoughts on the above implementations? ProcrastinatingReader (talk) 20:26, 4 September 2020 (UTC)
    @ProcrastinatingReader: preferably, this should be handled via a phab request, and not a js hack we have to run on every page load. — xaosflux Talk 21:55, 4 September 2020 (UTC)
    Xaosflux, is there some kind of file that adds the js on VE load only? Or is every page load the only option we have locally? ProcrastinatingReader (talk) 11:38, 5 September 2020 (UTC)
    Not that I'm aware of, and since it would be needed for ip users as well it would have to go in common.js - but server side (via a phab requests) perhaps those can be split, or tabbed, or at least reordered? — xaosflux Talk 22:20, 5 September 2020 (UTC)

Papyrus font?

I have had my signature set to display Papyrus font, which it did until today. I am guessing that Papyrus no longer displays for reasons other than Wikipedia settings (for lack of a better a term), but I am not technically adept, so I thought I would inquire here. I have "treat the above as wiki markup" checked. The markup code for my signature is:

<span style="font-family:Papyrus;font-size:14px;">[[User:Markworthen|Mark D Worthen PsyD]] [[User talk:Markworthen|(talk)]] [he/his/him]</span>

TIA - Mark D Worthen PsyD (talk) [he/his/him] 03:54, 6 September 2020 (UTC)

@Markworthen: That font makes it to the final HTML. Are you sure that your computer didn't lose the Papyrus font somehow? Can you see/use it on other websites or programs? Can you check what your signature here looks like from another computer? Jackmcbarn (talk) 04:02, 6 September 2020 (UTC)
I can see the signature above in the Papyrus font. It is working fine on my computer, in other words. – Jonesey95 (talk) 04:06, 6 September 2020 (UTC)
Good golly Miss Molly! You are absolutely correct. I quickly checked using Safari on my iPhone and, sure enough, my beautiful Papyrus signature appears in all its Egyptian elegance. ;^) Many thanks - Mark D Worthen PsyD (talk) [he/his/him] 04:08, 6 September 2020 (UTC)

Updating categories when template/module updated

So we have T132467 but it's 4 years old with no sign of progress. Looks like null edits to member pages fix this issue, though. Would it be allowable to just make up a bot that can make null-edits on all transclusions of a template, thus updating a category's members in hours rather than some arbitrary period of time when the job queue gets around to it? If so, would such a bot need a BRFA (since, well, the null edits wouldn't exactly show up anywhere?) ProcrastinatingReader (talk) 00:27, 6 September 2020 (UTC)

We had a null edit bot that seems to have disappeared. --Izno (talk) 02:19, 6 September 2020 (UTC)
Yes, Joe's Null Bot. Graham87 06:10, 6 September 2020 (UTC)
The best proposed approach I saw came up in the related T157670 and T159512. Legoktm was able to generate a report of "stale" pages and the last time they were edited. The proposal was to have a process or bot that null-edited pages based on that report, starting with the oldest pages, with the idea that no page would be stale longer than one month. – Jonesey95 (talk) 03:28, 6 September 2020 (UTC)

"{{circa}}"

Looking at the article on Mustafa Kemal Atatürk, I notice that the abbreviation "c" to indicate "circa" for his year of birth appears at both of its two instances - lede and infobox - to display with a small dot preceding the c and a microscopic squiggle following it. I doubt this was intentional, but would have no idea how to fix it. If other users don't see this, maybe it's just my monitor? But all other characters display normally. This is Chrome on Windows 10. Milkunderwood (talk) 01:42, 2 September 2020 (UTC)

@Milkunderwood: Are you describing the effects of the <abbr> tag? If so, then it's supposed to be like that, to indicate that you can hover over it to see what it's an abbreviation of. If not, can you upload a screenshot of exactly what you see? Also, does it happen on other pages, like here "c." on this one? Jackmcbarn (talk) 01:51, 2 September 2020 (UTC)
In your reply example here I see exactly the same thing I first saw: a small dot, smaller than a period, level with the bottom of the letter "c", and a very tiny squiggle following the letter, again level at the bottom of "c". Milkunderwood (talk) 02:04, 2 September 2020 (UTC)
@Milkunderwood: Can you post a screenshot? Jackmcbarn (talk) 02:08, 2 September 2020 (UTC)
I can, but I only know how to screenshot fullscreen. If I hover, it does show me "circa". So those extraneous-looking marks indicate "hover here"? Milkunderwood (talk) 02:14, 2 September 2020 (UTC)
I'm not sure a full screenshot on this page is a good idea, and I don't particularly want to put my entire computer settings out in public. I'll take a break and see if I can figure out how to do a small shot of just what I'm seeing. Milkunderwood (talk) 02:18, 2 September 2020 (UTC)
I have my shot of what I see, but nothing will let me paste it here, either as png or pdf. Is there a special trick to that part? Milkunderwood (talk) 02:58, 2 September 2020 (UTC)
The "squiggle" is an optical illusion caused by the juxtaposition of two entirely different things - the period denoting an abbreviation appearing just above the gap between two of the dots from the dotted underline. Try using your browser's zoom-in feature to see what I mean. As for the screenshot, as noted in the box displayed whilst editing, if a screenshot would help you to describe the problem, please follow the directions at Wikipedia:Screenshots of Wikipedia. --Redrose64 🌹 (talk) 19:42, 2 September 2020 (UTC)

On my monitor, I have a line break in the lede between the c. and 1881 - which is confusing enough as it is. Better to just spell out the word "circa" and be done with it. Milkunderwood (talk) 03:20, 2 September 2020 (UTC)

  • I also don't see the point of {{Circa}} template. On mobile and other contexts where there's no hover it's just a meaningless distraction and it's hard to see why we cannot use the full word. – Ammarpad (talk) 06:06, 2 September 2020 (UTC)
    MOS:CIRCA also suggests using just c. for circa. — GhostInTheMachine talk to me 07:13, 2 September 2020 (UTC)

@Milkunderwood: Is https://imgur.com/a/T7ubVt5 consistent with what you see? This is Chrome (version 84 on Windows 10) trying to be smart by interrupting the underline where it gets too close to letters. Because the underline is so short and spends most of its time almost touching the letter c, it is barely visible. This, that and the other (talk) 08:05, 2 September 2020 (UTC)

This sounds very much like Template talk:Abbr#Displaying strangely. --Redrose64 🌹 (talk) 19:08, 2 September 2020 (UTC)
Adding text-decoration-skip-ink: none; could fix this. Nardog (talk) 19:20, 2 September 2020 (UTC)
Only if the browser supports it. The text-decoration-skip-ink property has only been described in CSS Text Decoration Module Level 4 which was first published on 13 March 2018 and is still at the W3C Working Draft stage, hence is liable to be modified (or even completely removed) at any time. --Redrose64 🌹 (talk) 12:42, 3 September 2020 (UTC)
@Redrose64: It may not be appropriate as a site-wide style, but nowadays something like three quarters of web users use Blink-based browsers. Wouldn't it make sense to use it at least for this template as a tentative fix? Nardog (talk) 08:07, 6 September 2020 (UTC)

Many thanks for everyone's help and suggestions. Yes, https://imgur.com/a/T7ubVt5 is exactly what displays on my monitor. It's confusing by itself; but in the context of where it was placed in the lede at Mustafa Kemal Atatürk it was a real headscratcher. (But I still can't figure out how to post a screenshot here at Wikipedia. Different issue.) Milkunderwood (talk) 00:37, 3 September 2020 (UTC)

@Milkunderwood: As I noted in my post above, there are directions at Wikipedia:Screenshots of Wikipedia. --Redrose64 🌹 (talk) 12:44, 3 September 2020 (UTC)
Ah. I thought it might be a relatively straightforward and simple trick. But thanks. Milkunderwood (talk) 06:18, 4 September 2020 (UTC)

False edit conflict notice when I try to save - again

I used to get these but they went away. Now they are back. It's a pain because I have to then look at the history in a new tab to see if my edit has actually saved. This has happened twice today, both times my edit had saved. Doug Weller talk 15:56, 5 September 2020 (UTC)

I've had a couple of these in the past 24hrs. Literally only two (maybe three) out of 500+ edits, so I don't know what triggers it. Lugnuts Fire Walk with Me 07:10, 7 September 2020 (UTC)

Converting to data-sort-value

A bunch of the lists linked by {{CanadaRadio}} use old sorting tricks with hidden spans (especially List of radio stations in Quebec) and have done so for more than a decade. I tried to convert these to use data-sort-value, but I couldn't figure out the regex to do it properly. Could someone assist? Raymie (t • c) 23:43, 4 September 2020 (UTC)

@Raymie: The frequency column sorts correctly for me, as long as you don't mind the AM stations being sorted higher than the FM ones (because the AM freqs are in kHz and all above 600, making them numerically higher than the FM freqs that are all below 108 MHz). That could be solved by adding sortkeys to the relatively few AM stations, preferably with a simple template that takes the frequency in kHz as a parameter and spits out the data-sort-value (parm/1000) and display value. Note there's one AM station that has both AM and FM freqs crammed into the field, which should probably just be split into two rows. I didn't look too carefully for other similar issues. —[AlanM1 (talk)]— 09:39, 7 September 2020 (UTC)

Tech News: 2020-37

Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.

Changes later this week

  • Normally pages can be moved to a title that has no existing page yet or to a page that has only one revision, which is a redirect to the page to be moved. A new user right allows editors to move pages over one-revision pages that redirect to anywhere. [11]
  •   The new version of MediaWiki will be on test wikis and MediaWiki.org from 8 September. It will be on non-Wikipedia wikis and some Wikipedias from 9 September. It will be on all wikis from 10 September (calendar).
  •   All MediaWiki API modules will now use watchlist instead of watch. This was inconsistent before. [12]

Future changes

  • The Wikipedia Android app team might work on patrolling tools in the future. You can let them know what tools would be useful for you or for less experienced patrollers. See the page on mediawiki.org.
  • OTRS will be updated to a new version. This will probably take around two days. OTRS agents will not have access to the system during these days. Emails that come in during the update will be delivered when the update is done. The plan is to start around 08:00 UTC on 14 September. This could change. [13]
  • The Wikipedia Android app will send push notifications if users want them. This could help you see for example when someone wrote on your talk page or your edit was reverted. This will need Google Play Services to work. It will also be possible to get the app without Google Play Services but push notifications will not work. Google Play Services is also used to make the app work for Android 4.4 users. [14][15]
  •   Wikimedia code review could move to GitLab. It would be hosted on Wikimedia servers. You can take part in the consultation.
  •   Dropdown menus in the Vector skin use a .menu class. This will not work in the future. Scripts can use nav ul instead. .vectorTabs and .vectorMenu will also not work. Some scripts need to be updated. You can read more in Phabricator.

Tech news prepared by Tech News writers and posted by bot • Contribute • Translate • Get help • Give feedback • Subscribe or unsubscribe.

15:59, 7 September 2020 (UTC)

This is why we can't have nice things

For your viewing pleasure:

It goes without saying that one of these should be a redirect to the other, but these are actually different, apparently incompatible templates to make some sort of table row, both in use. I doubt it's a high priority, but I thought I'd just let technically-minded folks know in case someone feels like working on this. –Deacon Vorbis (carbon • videos) 22:34, 6 September 2020 (UTC)

Ideally the newer one should be merged to the older and functionality consolidated, but I see it's more developed than the older one. They also seem to be used in a slightly different way so probably it may not be easy merging, but I think it's worth doing. Templates should not differ based on trivial spelling difference like that. – Ammarpad (talk) 10:13, 8 September 2020 (UTC)
Any reason this can’t be sent to TfD? Editors there can figure out the technical feasibility of it. You’re not required to be able to merge yourself to nominate a discussion. ProcrastinatingReader (talk) 13:58, 8 September 2020 (UTC)

Toolforge problems?

Tracked in Phabricator
Task T262239

Oddities are happening today.

  • Most notable to me personally, is that when I bring up any Wikipedia article, there's usually a one-line read-out at the top of the page that tells me who created it, when, how many revisions, etc. That still holds for the older articles. But today when I'm working the CSD candidates, if it's probably a new creation, it reads this "Unable to fetch revision data. 0 pageviews (30 days)"
  • https://afdstats.toolforge.org/afdstats.py?name=Maile66&max=&startdate=&altname= has not been updating for me the last few hours. If I try to log in to https://wikitech.wikimedia.org/wiki/Portal:Toolforge, which I thought I was able to do, it tells me I'm giving the wrong user name and wrong password. Maybe I never could log in there - I don't remember. — Maile (talk) 22:46, 7 September 2020 (UTC)
The enwiki database replicas appear to be lagged by close to 21 hours. I've inquired on IRC, but predictably everyone is asleep. ST47 (talk) 05:23, 8 September 2020 (UTC)
I have asked on Phabricator. ST47 (talk) 05:40, 8 September 2020 (UTC)
The wikitech site is not a part of SUL. You need to create an account there to be able to log in. – SD0001 (talk) 06:46, 8 September 2020 (UTC)
  • This is a consequence of some maintenance that is taking place on the databases. The DBAs expect the lag to be there for a few more days. This will affect any tools which use the enwiki database replicas. ST47 (talk) 17:06, 8 September 2020 (UTC)

JavaScript help needed for file upload wizard

Following up from this discussion, we'd like to redesign the file upload wizard so that it asks right off the bat whether you're uploading a free or non-free file, and sends you to Commons if you're uploading a free one. This would be much better than the current ordering, where the many people who are trying to upload a free file don't realize they should go to Commons until the third step, at which point they'll have to re-enter a bunch of information (and some may just give up, costing us an image we'd want).

There seems to be some JavaScript code behind the wizard, though, which I don't know how to edit. Does anyone more tech-savvy want to take a stab at this? Sorry if this isn't quite the right forum, but it's an important outstanding task. {{u|Sdkb}} talk 21:39, 8 September 2020 (UTC)

@Sdkb: The JS, which is at MediaWiki:FileUploadWizard.js, reveals the relevant boxes (hidden at the start with display: none; in the page's wikitext) and fills in the contents of boxes with inputs and such. If the intention is just to reorder things, I think you could just do that by editing the contents of Wikipedia:File Upload Wizard. --Yair rand (talk) 21:53, 8 September 2020 (UTC)
Yair rand, it's basically just a reordering, yes. It'll be slightly more complicated, though, because the form currently has the actual "upload" button in step three for people who ignore the advice to go to Commons (why do we even allow that?), and that'll have to be moved to what's currently step 2.
Overall, the file upload wizard probably needs a somewhat fundamental redesign, but I don't know where to go to get a group together to work on it (other than WP:WikiProject Usability, which is typically me talking to myself). Maybe we could borrow the design of the Commons wizard, which is much better? But tackling the big ordering problem will hopefully be much simpler than a full redesign, and it'll be a good start. {{u|Sdkb}} talk 22:04, 8 September 2020 (UTC)

Need help Auto-minimize Draft tags for Visual Editor like we do with cleanup/AfD tags

[moved from proposal tab]

I’ve had bad user experiences (UX) while looking at Draft articles through the Visual Editor, which I use almost exclusively.

Instead of seeing the article itself, they’ll have massive templates that have to be scrolled through every time you want to see—let alone edit—the article.

In contrast, when I look at an article in main space that has a cleanup or AfD tag, it’s unobtrusive and minimized, and expands when you click on it. I think the various Draft tags should be likewise automatically minimized until they are clicked on. I think newer users especially would feel more comfortable continuing to improve the article if there wasn’t massive rejection tags atop them that have to be traversed every instance. Gleeanon409 (talk) 13:19, 18 August 2020 (UTC)

Gleeanon409, sounds reasonable to me. I'm not sure what the technical mechanisms at play are. {{u|Sdkb}} talk 19:06, 18 August 2020 (UTC)
Gleeanon409, could you please give me a couple of links to a "worst-case" draft, and a contrasting article with cleanup or AFD tags? I can file a bug report if the problem is in the VisualEditor instead of in the template design. Whatamidoing (WMF) (talk) 04:10, 20 August 2020 (UTC)
@Whatamidoing (WMF):,Dumb Show vs. Draft:Tor'i Brooks. Gleeanon409 (talk) 07:26, 20 August 2020 (UTC)

What needs to take place to make this happen? Gleeanon409 (talk) 12:46, 24 August 2020 (UTC)

@Gleeanon409, it looks like the problem is that the (very large) AFC box has several collapsed sections. When those are expanded, the whole thing is enormous. This can't be fixed in the visual editor, which treats all collapsisble content the same, and specifically uncollapses them (e.g., so you can make sure that you've added the correct navbox). But it might be possible to fix this in the template, using something like {{if preview}} to keep the content collapsed state. @Evad37, do you think that might work? Whatamidoing (WMF) (talk) 17:28, 24 August 2020 (UTC)
@Whatamidoing (WMF): Template:If preview doesn't work with visual editor - e.g. https://en.wikipedia.org/w/index.php?title=Template:If_preview/doc&veaction=edit shows the "when saved" values in the examples column, not the "when previewed" values. - Evad37 [talk] 01:16, 27 August 2020 (UTC)
Well, there goes that idea. Maybe we should take this to WP:VPT? Whatamidoing (WMF) (talk) 01:35, 2 September 2020 (UTC)
@Whatamidoing (WMF):, moved. Gleeanon409 (talk) 22:45, 8 September 2020 (UTC)

Autoheaders and infobox embedding

The guidance on embedding infoboxes (at Template:Infobox) says that, to have the title of an embedded infobox show up bold, it must be placed as a header. I want to use embedding on a set of infoboxes that have |autoheaders= enabled. Is there a way to specify that a specific header should always be displayed? Otherwise, the child will not display since it is assigned to a header field. Raymie (t • c) 23:47, 8 September 2020 (UTC)

Sitewide bug

What caused lists that have items that are listed horizontally to display incorrectly and have those items displayed vertically instead? A few minutes ago, the issue appeared on the Main Page and even in other articles. On the Main Page, bulleted lists had items that were displayed vertically. Even Template:COVID-19 pandemic data seems to have been affected, as all country names were shown in boldface, and the green box that was supposed to surround the documentation did not even display. LSGH (talk) (contributions) 07:30, 5 September 2020 (UTC)

Good question, I had the same problem. – Teratix ₵ 07:37, 5 September 2020 (UTC)
For about 15 minutes (~07:10 – 07:25 UTC), MediaWiki:Common.css failed to load. My Firefox console said
The resource from “https://en.wikipedia.org/w/load.php?lang=en&modules=site.styles&only=styles&skin=vector” was blocked due to MIME type (“text/javascript”) mismatch (X-Content-Type-Options: nosniff).
Nardog (talk) 07:38, 5 September 2020 (UTC)
I initially thought that the problem was with my browser, as it crashed and I needed to log in again. Other non-WMF sites were working perfectly fine. LSGH (talk) (contributions) 07:46, 5 September 2020 (UTC)

This is happening again right now. All https://en.wikipedia.org/w/load.php?lang=en&modules=site.styles&only=styles&skin=vector returns is: 

/* This file is the Web entry point for MediaWiki's ResourceLoader:
   <https://www.mediawiki.org/wiki/ResourceLoader>. In this request,
   no modules were requested. Max made me put this here. */

Nardog (talk) 14:04, 9 September 2020 (UTC)

...and it's gone. Nardog (talk) 14:06, 9 September 2020 (UTC)

Template:To do

  Resolved

Could a grown-up please publish the sandbox version of the template. — GhostInTheMachine talk to me 09:35, 9 September 2020 (UTC)

I've got this one. – Jonesey95 (talk) 16:00, 9 September 2020 (UTC)

This one weird trick will stop accidental logged-out editing

I use Firefox on Linux, with uBlock Origin. I use this rule to prevent accidental logged out editing, in the 2010 editor:

*##^#mw-anon-edit-warning ~ #editform .editOptions

This hides the "Publish" button and the edit summary field, and seems to stop Alt-Shift-S from working as well.

  • Anything I'm forgetting? That is, any way to accidentally save a page that this won't cover?
  • What works in your browser/OS/adblocker?
  • Should we make a list, and link to it from ... somewhere ... privacy-conscious users are likely to read?

This would seem more proactive than suppressing IPs after the fact. Suffusion of Yellow (talk) 23:06, 7 September 2020 (UTC)

The following actions will save your edit:
  • Alt+⇧ Shift+S at any time
  • ↵ Enter when the focus is on any of: the Subject/headline box (when making a new section); the edit summary box (when editing an existing section); checkbox "This is a minor edit"; checkbox "Watch this page"; Publish changes
  • Space when Publish changes has the focus
I don't know of other keyboard methods. --Redrose64 🌹 (talk) 23:18, 7 September 2020 (UTC)
If I go onto an edit page, and I've forgotten to log in, Wikipedia tells me I'm not logged in. I'm not sure I've understood what is the problem your solution addresses.Uporządnicki (talk) 00:07, 8 September 2020 (UTC)
Banner blindness. It's a common enough problem that it's mentioned specifically at WP:OSPOL. Suffusion of Yellow (talk) 00:47, 8 September 2020 (UTC)
Another option is to style the publish button to have distinct, non-default color. I use that myself, and the moment I see default blue color I know something is amiss. – Ammarpad (talk) 10:01, 8 September 2020 (UTC)
Or a simple solution is to change skin to anything but (spit) Vector, which will jarringly pop up when one gets logged out, since it's the default. Yngvadottir (talk) 19:34, 9 September 2020 (UTC)

Issues with the Wikipedia app?

Is anyone experiencing issues with the Wikipedia app? I have a Galaxy S10 and the articles are not loading. Only the images and the wiki-links show up. No text, no substance. -- Veggies (talk) 12:11, 9 September 2020 (UTC)

Apparently there is some instability on Parsoid, a component used for mobile app and the visual editor. Site Reliability Team is aware and working on it. Will update with a task or more info when I can.--JCrespo (WMF) (talk) 13:32, 9 September 2020 (UTC)

vis ed problems?

Anyone else finding Vis Ed is timing out? I'm having no trouble as long as I stick to source editing. —valereee (talk) 12:56, 9 September 2020 (UTC)

+1 Keeps on timing out. ~~ CAPTAIN MEDUSAtalk 13:16, 9 September 2020 (UTC)
screenshot ~~ CAPTAIN MEDUSAtalk 13:26, 9 September 2020 (UTC)
See above, apparently there is some instability on Parsoid, a component used for mobile app and the visual editor. Site Reliability Team is aware and working on it. Will update with a task or more info when I can.--JCrespo (WMF) (talk) 13:33, 9 September 2020 (UTC)
As far as I am told, visual editor and mobile app for the English Wikipedia should be back working normally, although some other languages and projects may still be affected. Please retry and report if it still doesn't work for you. --JCrespo (WMF) (talk) 13:38, 9 September 2020 (UTC)
The app still doesn't work for me. -- Veggies (talk) 16:52, 9 September 2020 (UTC)
Now it works. Mark as resolved for me. -- Veggies (talk) 21:30, 9 September 2020 (UTC)

How to test a patch for system JS

I think I've figured out how to fix phab:T102286 and phab:T260500 while fiddling with my browser, but I want to test it further before submitting it for code review, which I've never done before. Is there a handy way to do this without, say, setting up a server and installing MW from scratch? (Or if someone with experience can do it on my behalf, that would also be great, the code is here.) Nardog (talk) 14:33, 9 September 2020 (UTC)

You can use https://patchdemo.wmflabs.org/ tool. It will require you submitting the change to code review first though, but I think it's still easier than setting up mw dev environment from scratch. – Ammarpad (talk) 07:09, 10 September 2020 (UTC)

Citations button error

I just noticed the new "Citations" button on the edit page. Not sure exactly what it does but when I click it I get an error popup: "Error: Citations request failed". RedWolf (talk) 16:57, 8 September 2020 (UTC)

@RedWolf: can you show a screen shot? Also what platform are you on (web, mobile web, mobile), what skin are you using, which editor are you using? — xaosflux Talk 18:22, 8 September 2020 (UTC)
@Xaosflux: error pop-up Web (FF 80), MonoBook, using the standard editor (not visual editor). RedWolf (talk) 18:57, 8 September 2020 (UTC)
RedWolf, do you have this gadget enabled ? Wikipedia:Citation_expander —TheDJ (talk • contribs) 20:32, 8 September 2020 (UTC)
@Xaosflux: Yes, citation expander gadget enabled. RedWolf (talk) 20:50, 8 September 2020 (UTC)
It's just being used a lot. I think it should be separated from Citation bot. Abductive (reasoning) 08:46, 10 September 2020 (UTC)

Image replacement failure - advice sought

Seem to have ended up with a different image in the file to the one displaying on the page (but it is OK on my iPhone)Horatius At The Bridge (talk) 07:09, 10 September 2020 (UTC)

@Horatius At The Bridge: See Wikipedia:Bypass your cache. Nardog (talk) 08:36, 10 September 2020 (UTC)
Magic   Many thanks Nardog Horatius At The Bridge (talk) 08:51, 10 September 2020 (UTC)

Lack of a "Smart toggles/tabs" JavaScript modifying internal custom mw-collapsible behavior

Several months have passed and the Medical cases chart template still suffers from a limitation of the mw-collapsible plugin. What is wanted is that toggles dynamically affect (based on their expanded, partial or collapsed states) their linked collapsible elements and sibling toggles, instead of the current "blind" toggling which creates the intersection of sets mess described in that talk page. I have checked all MediaWiki:*.js titles in MediaWiki and Wikipedia and found no JS which implements a similar functionality. Since such feature sounds so useful and basic/"core", I wondered if its absence was due to some kind of policy against JS which improves the visualization of content in specific templates. If there was nothing against it and safety guidelines were followed, would a .js page creation request from a non-admin possibly be accepted? Alexiscoutinho (talk) 00:13, 2 September 2020 (UTC)

@Alexiscoutinho: so basically the problem is that there is not a way for a page or template to "call a javascript" (at least not until phab:T8883 or phab:T63007 ever get done) - so the only way to have a local script run is to have it run on every page load (several ways to do that - but all have that affect). The best way to get a new table handler would be to place a enhancement request on phab. — xaosflux Talk 02:28, 2 September 2020 (UTC)
@Xaosflux: I agree, but in the meantime, as a workaround, the Load JS and CSS by URL snippet could be used. If the script gets good and robust enough, then it would make sense to merge it as an extension into the plugin itself. I just wanted to be sure that putting work on a JS that alters the event handlers, data parameter and so on of custom collapsible content wouldn't go to waste by an obscure policy... By the way, do you know where (in which source code) is makeCollapsible initially called to configure all mw-collapsible elements? I know Common.js calls it for a deprecated collection, so there should be at least another call buried in MediaWiki core. Also, are the hook handlers run in the exact order in which they are added/registered to the wikipage.content event? Alexiscoutinho (talk) 03:51, 2 September 2020 (UTC)
Alexiscoutinho, the mw-collapsible code is initiated by the mediawiki.page.ready core module. The code itself is in the jQuery.makeCollapsible core module. A postprocessing hook (which implements autocollapse and innercollpase is in MediaWiki:Common.js at the mwCollapsibleSetup function. "are hook handlers run in the exact order in which they are added/registered to the wikipage.content event?" Yes, but that registration order is asynchronous so there is no guaranteed order in the registration fase. —TheDJ (talk • contribs) 09:11, 2 September 2020 (UTC)
@TheDJ: Thank you! I didn't think that there was a github behind the API documentation. Alexiscoutinho (talk) 12:58, 2 September 2020 (UTC)
Alexiscoutinho, the GitHub is a mirror of our gerrit repos. But they are easier to browse/search at times. —TheDJ (talk • contribs) 13:22, 2 September 2020 (UTC)
TheDJ, is it possible to remove the handler that calls makeCollapsible in ready.js before it is first run? Not only is the handler outside my scope but it is an anonymous function... Removing it would help my testing since I would be able to process in my way and step-by-step the collapsible content without it already starting pre-processed by the default makeCollapsible. Alexiscoutinho (talk) 23:12, 2 September 2020 (UTC)
Alexiscoutinho, no this happens pretty early in page initialization, earlier than you can reach from a userscript. —TheDJ (talk • contribs) 07:47, 3 September 2020 (UTC)
TheDJ, is it bad to store information that is internally used by the plugin in jQuery's data object? I.e. the list of collapsibles and sibling toggles affected by a custom toggle. Should I only store information useful for an external user, like utility functions? Alexiscoutinho (talk) 22:52, 5 September 2020 (UTC)
Alexiscoutinho, no, that is not bad in itself. The collapsible module does it actually. However, in general you would avoid it, as it is faster to store and retrieve information in function scope of an instance of code that is being used somewhere. But the make collapsible plugin is a bit old, so it probably doesn't adhere fully to such criteria. An example of scoped storage would be the cache variable in the table initialiser of the tablesorter plugin. —TheDJ (talk • contribs) 07:48, 8 September 2020 (UTC)
@TheDJ: Also, should event handlers be somehow implemented to run in parallel as far as possible across Wikipedia or at least in the makeCollapsible plugin? Alexiscoutinho (talk) 14:41, 6 September 2020 (UTC)
No. experience has shown that this creates a huge processing peak, which is annoying for many users. Especially when dealing with HTML of content, simply process sequential from top to bottom. —TheDJ (talk • contribs) 07:48, 8 September 2020 (UTC)
TheDJ, I followed your advice and finished my changes to makeCollapsible. My code is here and it also contains an "unmaker" to bridge the behavior change. I also created a task (phab:T262622) following Xaosflux's suggestion. Alexiscoutinho (talk) 23:02, 10 September 2020 (UTC)

Can't the radio-button thing seen in infoboxes for places (MediaWiki:Gadget-switcher.js) be employed to achieve this? Nardog (talk) 07:32, 2 September 2020 (UTC)

I don't know which radio-button you are talking about. Could you show a template where it is used? However, it wouldn't make sense to add any radio-button to the Medical cases chart template. Furthermore, the modified toggle event handling would have to be implemented somewhere and MediaWiki:Gadget-switcher.js doesn't do that. Alexiscoutinho (talk) 12:58, 2 September 2020 (UTC)
@Alexiscoutinho: Just look up a place, e.g. Warsaw. Nardog (talk) 16:16, 2 September 2020 (UTC)

@Alexiscoutinho: Check this out (source/CSS):

Tab 1
Content 1
Tab 2

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

Nardog (talk) 03:32, 3 September 2020 (UTC)

@Nardog: Cool, but it seems just too limited. In the current state, that demo does not cycle between on and off if just one button is pressed consecutively. The "Show all" button doesn't highlight all tabs. Would initial collapsed states work? And last but not least, how would you go around internally representing partial states for the buttons (which would look like dashed outlines)? If you can show that your method works for a table with 2 buttons and 4 rows, where the 1st button covers/toggles the first 3 rows and the 2nd button covers the last 3 rows, and the third row is initially collapsed, and the button states (based on how many of their "children" are being displayed) are always accurately shown by their outline, I would be amazed and could even use it! Alexiscoutinho (talk) 04:40, 3 September 2020 (UTC)

Search busy

Tracked in Phabricator
Task T262694

Any searches I run (for things that don't explicitly match an article title) give me the error: "An error has occurred while searching: Search is currently too busy. Please try again later." on at least 9/10 attempts. I just noticed this a few minutes ago. Is this really something that might happen and should be ignored, or does this possibly indicate a problem? –Deacon Vorbis (carbon • videos) 16:56, 11 September 2020 (UTC)

Same for me. Kaldari (talk) 17:16, 11 September 2020 (UTC)

How were 50,000 bytes lost in 2001?

I decided to look at the earliest versions of Empire State Building. Look at the number of bytes changed in the third version.— Vchimpanzee • talk • contributions • 20:11, 11 September 2020 (UTC)

There was some weird glitch back then. At one point every edit to September 11 attacks seems to have resulted in a loss of over 100,000 bytes.— Vchimpanzee • talk • contributions • 20:21, 11 September 2020 (UTC)
Both very likely a result of Graham87's imports from nost:, the same way the byte counts get confused when you do a history merge or split. —Cryptic 20:29, 11 September 2020 (UTC)
Yes, this is T114806, which was not fixed retroactively. Also see T38976. Graham87 08:33, 12 September 2020 (UTC)

New bot task

I've recently written two scripts, User:Ritchie333/wirblps.py and User:Ritchie333/unlinkedblps.py, which populate User:Ritchie333/Unreferenced women BLPs and User:Ritchie333/Unreferenced and unlinked BLPs. These are used by Wikipedia:WikiProject Women in Red to track and fix various BLPs without sources.

In the past, Galobtter has integrated some of my scripts into his bot, but he's not available right now. I could set my own bot up, but it just seems like a lot of hassle for something that is going to run at most every hour, and write to a maximum of two pages, both in my userspace. Can anyone else help? Ritchie333 (talk) (cont) 10:35, 11 September 2020 (UTC)

A better way to establish the gender is to use the linked wikidata item rather than look for gendered categories which are unlikely to exist in many cases. Looking at the ORES articletopic classification is another way. You should put content=True parameter in calls to CategorizedPageGenerator which would speed up the script a great deal (and simultaneously makes it mw:API:Etiquette-compliant - fetching a number of pages together in a single API call is preferable). Also, use {{annotated link}} template rather than dumping a rawlist, so that the short descriptions are visible.
I can commit to this after a few days when I get time, if no one beats me to it. – SD0001 (talk) 14:00, 11 September 2020 (UTC)
@Ritchie333: Did you try WP:BON and/or WP:BOTREQ? --Redrose64 🌹 (talk) 15:50, 12 September 2020 (UTC)

LaTeX rendering problem

Can anyone find a fix for the ridiculously tall square-root and brackets in this:

 

It appears that the height is being calculated as if the numerator and denominator inside the square root are the same height, which clearly they are not. More sensible results if the p/n element in the numerator (the tallest bit) isn't there,

 

Thanks SpinningSpark 13:40, 9 September 2020 (UTC)

Never mind, I worked it out myself, the final bar was supposed to be outside the sqrt
 
but I'm still not sure why that was messing up the heights. SpinningSpark 14:17, 9 September 2020 (UTC)
I think this would be a job for \smash ...which is yet another bit of TeX that's not supported. Otherwise, it just seems to choke a little when the stretchable delimiter sizes get too large and it's desperately trying to keep trailing text centered with the fraction base line. (Note it looks fine if you take out the kappa). I've certainly seen this happen in other sorts of formulas, and you can try to rewrite the formula so it's not an issue (ec, already done apparently), but otherwise, there's not much we can do from this end. Maybe Debenben might know a little more? –Deacon Vorbis (carbon • videos) 14:23, 9 September 2020 (UTC)
Actually, in this case it was due to invalid formatting. The bracket was opened outside the square root, but closed inside it. It just was not obvious because the opening bracket was an "invisible" one. I'm surprised it rendered at all. SpinningSpark 17:34, 9 September 2020 (UTC)
I am not sure why the height calculation is that way but it is correct because in LaTeX it is the same. Things like \smash and \llap would be nice to have, those TeX commands work fine in MathJax. @Physikerwelt: --Debenben (talk) 16:39, 12 September 2020 (UTC)

Draft pages turning up in a Google search

Question: I recently speedy deleted several drafts per G3 as hoaxes. In doing a BEFORE search at Google I was surprised to find that the drafts turned up in a Google search. I thought drafts were not supposed to be indexed? -- MelanieN (talk) 20:55, 12 September 2020 (UTC)

For specific examples: two of the drafts that turned up in a Google search were Draft:Discovery (Christina Aguilera album) and Draft:Dayanithi. -- MelanieN (talk) 20:59, 12 September 2020 (UTC)
How old were the drafts? Adam9007 (talk) 21:02, 12 September 2020 (UTC)
May 2020 and October 2018. -- MelanieN (talk) 22:24, 12 September 2020 (UTC)
The latter was speedied in April 2019, and not recreated until 9 September 2020.
I don't suppose there's any chance what you saw was a google hit of a transclusion of one of these drafts on another page, rather than the draft itself? Or maybe a redirect from an indexed namespace to the draft? Not that I see any edits like those in the contribs or deleted contribs of any of the editors of those pages (except perhaps the speedy tagger, who I didn't check); I'm just kind of grasping at straws at this point. —Cryptic 22:30, 12 September 2020 (UTC)
This came up a few months ago. Google has been indexing the page titles but the page text shouldn't be displaying in results. Is it? --Izno (talk) 01:06, 13 September 2020 (UTC)

Side box width seems slightly off

 – {{u|Sdkb}} talk 02:13, 14 September 2020 (UTC)

I noticed here that the Wikisource template (which ultimately relies on Template:Side box) is slightly wider than the normal thumb images. Would it be possible to adjust the width to bring them into alignment? {{u|Sdkb}} talk 19:33, 12 September 2020 (UTC)

The display size of thumb images is subject to preferences (and presumably site-wide settings). The size of side boxes, in this case 238px, seems to be determined by some classes, so it seems impractical and probably futile to harmonize them. -- Michael Bednarek (talk) 05:03, 13 September 2020 (UTC)
Michael Bednarek, it's true that users can change their preferred thumb image display size, but 99.9% of readers never do so, so it's of very little pertinence. I'm not fully familiar with classes; could you clarify what'd make it impractical to change them? {{u|Sdkb}} talk 01:36, 14 September 2020 (UTC)
Like you, I'm unfamiliar with the way classes in Wikipedia operate. Where are they defined? Who defined them? Is there a process to get community input in their creation or modification? That's why I called it impractical. Maybe WP:VPT is a better venue to raise this. -- Michael Bednarek (talk) 01:45, 14 September 2020 (UTC)
Sounds good; I'll move this there. {{u|Sdkb}} talk 02:11, 14 September 2020 (UTC)
The class that controls the size of the Wikisource box is .mbox-small. It's defined in the common CSS file (line 819), and has a default width of 238px (contrary to the default thumbnail width of 220px). Since the way thumbnails are displayed also change per skin, the template can be smaller than thumbnails in other skins (e.g. Timeless (test)). As far as I am aware of though, no magic word provides the user's preferred thumbnail width, which would be needed in order to set the size of {{Side box}} to the user's preferred size (which would be done by setting the template width to the preferred thumbnail size through the style attribute, like how thumbnails do now.) That unfortunately means we'll have to stick with the differences in width, unless we can somehow get the width of the thumbnails. This, of course, doesn't factor in some accessibility issues that doing this might have, which I can't comment on since I'm not well-versed in accessibility-focused web design. Chlod (say hi!) 02:51, 14 September 2020 (UTC)

Tech News: 2020-38

Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.

Problems

  • The Wikipedia apps briefly showed pages without CSS last week. This meant they looked wrong. It was quickly fixed but cached pages without CSS were shown for a few hours. [16][17]

Changes later this week

  •   The new version of MediaWiki will be on test wikis and MediaWiki.org from 15 September. It will be on non-Wikipedia wikis and some Wikipedias from 16 September. It will be on all wikis from 17 September (calendar).

Tech news prepared by Tech News writers and posted by bot • Contribute • Translate • Get help • Give feedback • Subscribe or unsubscribe.

16:18, 14 September 2020 (UTC)

Help with regex

Hey there! :)

Is someone free to help me a bit to combine two find and replace regular expressions into one? I need that for a bot I'm using in a wiki but I'm a bit new to regexes. To not over clutter this section, we can hold the conversation at a user talk page (either mine's or another's) so just tell me below and I'll write to you (or you write to me). Thank you in advance! :) - Klein Muçi (talk) 09:42, 13 September 2020 (UTC)

Klein Muçi: You're honestly better off just saying what you want to do here so that you get more eyes on it. No one worries about clutter; just put it up here.--Jorm (talk) 15:25, 13 September 2020 (UTC)
@Jorm: okay then, if you insist... :P
So I have a table like this in Notepad++:
Klein Muçi, for the future, it can be handy to make use of something like: https://regex101.com to build your regular expressions and to easily see what parts match what content for which reason. —TheDJ (talk • contribs) 07:08, 14 September 2020 (UTC)
Extended content
aa: Afar
ab: Abkhaz
ae: Avesta
af: Afrika
ak: Akan
(many more entries)
Firstly, I search for:
^.*: (with space at the end)
And replace it with:
"\\|\\s*language\\s*=\\s*
That gives me:
"\|\s*language\s*=\s*Afar
"\|\s*language\s*=\s*Abkhaz
"\|\s*language\s*=\s*Avesta
"\|\s*language\s*=\s*Afrika
"\|\s*language\s*=\s*Akan
Secondly, I search for:
'^.*
And replace it with:
($0)\\b" "|language=
That gives me:
"\|\s*language\s*=\s*Afar\b" "|language=
"\|\s*language\s*=\s*Abkhaz\b" "|language=
"\|\s*language\s*=\s*Avesta\b" "|language=
"\|\s*language\s*=\s*Afrika\b" "|language=
"\|\s*language\s*=\s*Akan\b" "|language=

I want to be able to go from:

aa: Afar
ab: Abkhaz
ae: Avesta
af: Afrika
ak: Akan
(many more entries)

to:

"\|\s*language\s*=\s*Afar\b" "|language=
"\|\s*language\s*=\s*Abkhaz\b" "|language=
"\|\s*language\s*=\s*Avesta\b" "|language=
"\|\s*language\s*=\s*Afrika\b" "|language=
"\|\s*language\s*=\s*Akan\b" "|language=
(many more entries)

in one step instead of two. I'm assuming that's possible. How do I do that? - Klein Muçi (talk) 16:15, 13 September 2020 (UTC)

I think I'm confused. You have a file with lines "xx: Yyyyyy"; and you want them into an escaped format? Probably better to show me one line that you have and then line you want it to look like. The way I'm understanding your example is trivial and shouldn't require two regexes to make it.--Jorm (talk) 16:23, 13 September 2020 (UTC)
This?
Find: [a-z]{2,3}:\s*([^\n\r]+)
Replace: "\\|\\s*language\\s*=\\s*$1\\b" "|language=
doubled \ in the replace required in my copy of notepad++ may not be required in awb or some other regex tool.
—Trappist the monk (talk) 16:26, 13 September 2020 (UTC)
The F&R regex Trappist gave above works perfectly. Can you explain to me the specific function of its components? - Klein Muçi (talk) 16:53, 13 September 2020 (UTC)
Klein Muçi, "any alpha character ([a-z]), 2 or 3 times ({2,3}), followed by a colon (:), and then 0 or more whitespace characters (\s*), and then gruop the word to the end of the line into a variable (take everything from that point on to the end of line". Putting things inside of parens creates a variable to be used next, and they are addressed like $1, $2. Jorm (talk) 17:13, 13 September 2020 (UTC)
TheDJ I tried that page before coming here but I just couldn't create the logic for the expression that I was looking for. I'd really like it if we had a page that creates regex-es for us according to our description but, alas, my wish won't come true anytime soon. Jorm why "2 or 3 times"? - Klein Muçi (talk) 10:12, 14 September 2020 (UTC)
{2,3} means 2 or 3 times. Jorm said as much. You may find this detailed cheatsheet on regex syntax useful. – SD0001 (talk) 19:11, 14 September 2020 (UTC)
Assuming the characters before the colon in your example are language codes, the commonly used ISO 639-1 and 639-2 language codes use two or three letters. If your data has different characteristics, then you can adjust the count accordingly. isaacl (talk) 23:55, 14 September 2020 (UTC)
If you are interested in the opposite—turning a regular expression into a text description of its components—I suggest looking at the regular expressions 101 site. (There are other ones out there.) isaacl (talk) 00:00, 15 September 2020 (UTC)

Partial blocks for API calls

This is kind of expanding on the idea on a blacklist for RedWarn and other automated tools. But what we know is clear: you must use the Wikipedia API within Wikipedia policies or risk being blocked from editing.

Because it is easy to have a blacklist circumvented with modifications of a code or by using a different tool to achieve disruption, I suggest implementing something that blocks POST API calls only (maybe as part of m:Partial blocks). That way, the partially blocked editor can continue to edit Wikipedia with our web forms, etc., but cannot use any external tools or programs or gadgets to make edits, forcing them to go slow. Aasim 04:36, 15 September 2020 (UTC)

You'll be surprised to find out that VisualEditor uses the same API that all tools use. There is no private backend for on-site tools like VisualEditor. Unless we have each tool identify themselves in the API call, this is impractical. Even if we had each tool identify themselves and block depending on the identification, anyone can just circumvent that by using whatever VisualEditor uses in order to falsely identify yourself as a user of the VisualEditor. And if you only wanted to block a user from a specific tool, you could have just used that tool's own blacklist. Even if we were able to do API-level blocks (assuming that it's possible), anyone can just use a custom script to automate the process of editing an article. And of course, we can't even such achieve extreme levels of disruption unless the given user can bypass edit filters and at the same time evade blocks, since they'd most likely be blocked on sight for disruptive editing. Chlod (say hi!) 04:50, 15 September 2020 (UTC)
Awesome Aasim, just block people who don't abide by the rules. —TheDJ (talk • contribs) 08:34, 15 September 2020 (UTC)

Subcategory count

How do I retrieve the number of pages in a category and all of its subcategories? Neither {{Category count}} nor {{PAGESINCATEGORY}} seem up to the task, since they each count subcategories as only one item within the category. {{u|Sdkb}} talk 05:03, 14 September 2020 (UTC)

PetScan may be able to help, as long as the count is not too high. This search shows the templates in Category:CS1 errors and its subcategories, three deep. You may be able to modify it to get what you want. – Jonesey95 (talk) 05:13, 14 September 2020 (UTC)
Jonesey95, I'm looking to retrieve the total number of lists on Wikipedia for the WP:Featured lists intro (analogous to the WP:FA intro), so a one-time PetScan result unfortunately won't help since it doesn't update. {{u|Sdkb}} talk 05:34, 14 September 2020 (UTC)
There's no way to get a dynamically updating figure (save for a bot updating the figure on the fly using the result from petscan). – SD0001 (talk) 08:26, 14 September 2020 (UTC)
SD0001, hmm, oh well. Do we even know how to get a one-time count of the number of lists on Wikipedia? The PetScan query I made starting from Category:Lists didn't work since there must be a "leak" somewhere in the tree. And WikiProject Lists claims far fewer featured lists than WP:FL, so there must be a bunch of untagged pages (which is a little concerning for featured content). {{u|Sdkb}} talk 19:39, 14 September 2020 (UTC)
The {{Category count}} template uses {{PAGESINCATEGORY}} to obtain its figures. {{PAGESINCATEGORY}} is not a template but a parser function, and cannot go more than one level deep. As far as plumbing a category's depth is concerned, you need some kind of limit in case of category loops. The tightest possible loop is one level, and there are normally just four categories that are immediate members of themselves - any more than four indicates an error somewhere. The four are:
The first two are intentional, the second two are because of the way that they're constructed. --Redrose64 🌹 (talk) 20:47, 14 September 2020 (UTC)
Actually it turns out there were 12 more such cats containing themselves !   Fixed – SD0001 (talk) 04:55, 15 September 2020 (UTC)
@SD0001: I should have mentioned that they're reported (roughly) weekly at Wikipedia:Database reports/Self-categorized categories. --Redrose64 🌹 (talk) 12:33, 15 September 2020 (UTC)
@Redrose64: Cool. BTW you may be interested in the discussion at Wikipedia:Village_pump_(idea_lab)#Killing_cats. – SD0001 (talk) 14:28, 15 September 2020 (UTC)

Empty footnote apparently from a wikidata item

I'm sure the solution to this is simple, but it's outside of the info I have in my head. There's an empty footnote (currently number 5) in the Sariaya article. It is being caused by the line in the infobox reading: | elevation_footnotes = {{PH wikidata|elevation_footnotes}}. Wtmitchell (talk) (earlier Boracay Bill) 16:02, 15 September 2020 (UTC)

I suppose this line is the problem in {{PH_wikidata}}:
| elevation_footnotes = <ref>{{safesubst:#invoke:wd|reference|raw|P2044}}</ref>
When there is no reference it still passes the empty ref tags. It needs to put the reference in an if statement checking the reference. —  Jts1882 | talk  16:35, 15 September 2020 (UTC)
It seems as if you're saying that there is a problem either in {{infobox settlement}} or in specifying the content for elevation_footnotes as that infobox is used in the Sariaya article. I looked at the docs for that infobox and didn't see an answer; there's a commont on that parameter in the source for that infobox template which says, "for references: use <ref> tags", but I'm not sure what that is trying to tell me. Could you clarify this enough for me to figure out whether I need to follow this up on the article talk page or the talk page for the infobox template? Wtmitchell (talk) (earlier Boracay Bill) 22:30, 15 September 2020 (UTC)
(added) On further thought, it seems as if there is a problem with putting a {{wd}} invocation inside of a <Ref> because it might return an empty string -- which is sort-of what you said. However, from the viewpoint of the article editor, that is too far down in the weeds to be expected that it will be understood. If I'm right about that, I still don't see where I ought to follow this up. Perhaps there is a template somewhere which checks a string and, if it is empty, replaces it with an appropriate {{error}} (???). Wtmitchell (talk) (earlier Boracay Bill) 23:11, 15 September 2020 (UTC)
(added) Or maybe there needs to be a template named something like {{footnote|group=group|name=name|content}} which generates a <Ref> if content is not empty and does nothing if it is empty -- but that should be discussed someplace more appropriate than here (where?). Wtmitchell (talk) (earlier Boracay Bill) 23:41, 15 September 2020 (UTC)
{{PH wikidata}} should check for content in that Wikidata item before displaying <ref>...</ref> tags. I recommend opening a discussion at Template talk:PH wikidata. (Edited to add: I have started a discussion there.) – Jonesey95 (talk) 00:07, 16 September 2020 (UTC)
Sorry for not being clearer . What I meant, as Jonesey says, is that the Wikidata value needs checking before adding the reference tags. Something like:
| elevation_footnotes = {{#if:{{#invoke:wd|reference|raw|P2044}}
                        |<ref>{{safesubst:#invoke:wd|reference|raw|P2044}}</ref>
                        |}}
The spaces might need removing. —  Jts1882 | talk  07:39, 16 September 2020 (UTC)
  Done It fixes the problem at Sariaya and other places like Manilla, but I can't find an example with an elevation to check it is working when Wikidata has a value. —  Jts1882 | talk  08:15, 16 September 2020 (UTC)

Mobile site stripping page elements

I suppose I should walk through this. Robert le diable leads off with an image of Act III in its infobox. However, if viewed on the mobile site, the infobox - and image - disappears completely. What's going on, and if it's possible for the mobile site to remove elements, is it possible to provide an alternative (such as a thumbnailed image) for on the mobile site? Adam Cuerden (talk)Has about 7.5% of all FPs 22:13, 16 September 2020 (UTC)

It's not an infobox. It's a box that uses {{Composer navbox}}, which in turn uses {{sidebar with collapsible lists}}, which, according to its documentation does not display in the mobile view of Wikipedia. I'm pretty sure that all normal sidebars and navboxes are hidden in the mobile view. Someone else will have to answer the "display alternative image on mobile" question. – Jonesey95 (talk) 22:28, 16 September 2020 (UTC)
That is because the infobox is actually a sidebar, which are known to be removed on mobile. I am somewhat skeptical that fact should change since it is predominantly about the links provided, so that means that the image should perhaps be moved out of the sidebar, as images for sidebars are intended to be decorative in most cases. --Izno (talk) 22:29, 16 September 2020 (UTC)
The trouble with that is the concept of "lead image" conflicts with the idea that an infobox should be the first thing on a page. It's pretty darn common for - as an example - a short species article to put its only picture in the infobox, e.g. Xenops. The images are not being used solely for decoration, and I don't think that's ever been true. Adam Cuerden (talk)Has about 7.5% of all FPs 00:45, 17 September 2020 (UTC)
My point is that this case is not an infobox, it is a sidebar. If you want to image to show, you should take it out of the sidebar and make it a lead image. Most sidebars have decorative images. Most infoboxes do not. (Perhaps a discussion of interest.) --Izno (talk) 01:42, 17 September 2020 (UTC)

Setting ILL and EL colors

I'm trying to develop a dark mode color palette for the vector skin and am having trouble setting the color for visited interwiki and external links. Also, I'm having trouble changing the color palette for the text in the sidebar. (The WIP is available at User:LaundryPizza03/nightvector.css. The colors for interwiki and external links, visited and unvisited, should be the same as for internal links.) –LaundryPizza03 (dc̄) 08:28, 14 September 2020 (UTC)

Another question that was redirected from the idea lab: The dark mode has white text on a black background; the colors of links are also changed for legibility. I noticed that various templates with a colored background make the text illegible. I was wondering if a module could be implemented to adjust the template's colors to match the user's default text color, or if the text in such templates should default to black. Alternatively, is there a way to work around this issue? –LaundryPizza03 (dc̄) 05:35, 17 September 2020 (UTC)

Edit count

Tracked in Phabricator
Task T262239

Apologies if this is not the appropriate place to ask about Edit count. I've looked around and couldn't find a better place to do so. As far as I can see, Edit count has not been working for at least 24 hours. Is that so? Thanks. Misha Wolf (talk) 19:19, 8 September 2020 (UTC)

Misha Wolf, It's working for me. What happens when you try it? -- RoySmith (talk) 19:23, 8 September 2020 (UTC)
The field "Edits in the past 24 hours" shows "0", though I have made 5 edits in that period. Other stats are wrong too. For example, page https://xtools.wmflabs.org/articleinfo/en.wikipedia.org/Talk:ISO_4217 states that my "Latest edit" of Talk:ISO_4217 took place on/at 2020-09-04 20:28. In reality, I have made 7 edits to that page since that date/time. Thanks. Misha Wolf (talk) 19:32, 8 September 2020 (UTC)
As mentioned just above, this is because the toolforge database replicas are lagged due to ongoing maintenance. ST47 (talk) 19:39, 8 September 2020 (UTC)
Thanks! Misha Wolf (talk) 19:48, 8 September 2020 (UTC)
Basically, Wikipedia:Village pump (technical)/Archive 183#Toolserver replication lag for enwiki is now over 36 hours is happening again. --Redrose64 🌹 (talk) 22:10, 8 September 2020 (UTC)
Shame this is becoming so unreliable again. It used to run well until recently which was such an improvement to the way it ran (or did not run) a few years back. Have 2 day old edits at World number 1 ranked male tennis players which have not shown up yet, but also no indication there is a replication lag.--Wolbo (talk) 15:55, 9 September 2020 (UTC)
It's recovering, but is still not right. As with last time, there was one day when only one person managed to score an increase - congratulations Flyer22 Frozen (talk · contribs) for making two edits in the 24 hours ending 04:06, 11 September 2020. --Redrose64 🌹 (talk) 15:55, 12 September 2020 (UTC)
I did this twice, did I? Interesting...considering the many editors who spend significantly more time on Wikipedia than I do. Except for days like today (which find me on for longer), Wikipedia often gets only four hours out of me per day. But then again, I do intermittently use WP:Huggle, which inflates my edit count. And before that, I used WP:STiki (before it stopped working with this computer I use), which inflated my edit count a lot more. Flyer22 Frozen (talk) 01:56, 13 September 2020 (UTC)
Still a full day lag in the edit count.--Wolbo (talk) 21:30, 15 September 2020 (UTC)
I still notice the lag on just about any new page created. I have a script that normally gives me a read-out at the top of any page that tells me who created the page, how many editors, how many page views. etc. But if the page was created in the last day or two, it reads "Unable to fetch revision data. 0 pageviews (30 days)". — Maile (talk) 22:48, 15 September 2020 (UTC)
@Maile66: Did you check phab:T262239 - or even the replag itself - before posting here? --Redrose64 🌹 (talk) 18:08, 16 September 2020 (UTC)
... and your point is ... what? I was commenting on the couple of posts above mine. — Maile (talk) 01:19, 17 September 2020 (UTC)
My point is that not only is it a known problem, it is out of our hands and there is no point in making further posts here unless the situation is improving or even resolved. Every report, bot or script that obtains its data from Toolforge is affected. As I write this, the replag exceeds 69 hours with no prospect of immediate improvement; we at VPT can do nothing about it, the place to chase it up is the phab ticket. --Redrose64 🌹 (talk) 08:40, 17 September 2020 (UTC)

NOINDEX not working

I've added "__NOINDEX__" to my metawiki page to prevent Google being spammed with user pages when my name is searched, yet these pages still appear on Google; how do I get the tag to work? Sidenote: even random talk pages of mine show up; shouldn't non-mainspace be noindexed anyway?  Nixinova   T   C   06:59, 15 September 2020 (UTC)

Nixinova, you have to wait until Google revisits the page and this can take a while. Also yes, you have to noindex every page you want not listed separately. —TheDJ (talk • contribs) 08:32, 15 September 2020 (UTC)
Pages that have had the tag for like a year still show up in Google, however. Is there any way to fix that?  Nixinova   T   C   08:49, 15 September 2020 (UTC)
Nixinova, call google :D —TheDJ (talk • contribs) 09:46, 15 September 2020 (UTC)
TheDJ, BTW, the way to "call google" is via their webmaster tools. -- RoySmith (talk) 15:05, 15 September 2020 (UTC)
The webmaster tools can be used if you control the website; I'm not sure you can request a reindexing for a page on a site you don't control, though. isaacl (talk) 20:17, 15 September 2020 (UTC)
Namespace indexing parameters differ by project, if you want to propose a change to a namespace indexing policy on metawiki you will need to start an RfC there. — xaosflux Talk 14:47, 15 September 2020 (UTC)
Maybe related: phab:T90475 – Majavah talk · edits 09:41, 17 September 2020 (UTC)

wiki data bans iranian ip when signed in

when im logged in to wikidata why can i only do edis in wikipedia android app Baratiiman (talk) 11:41, 17 September 2020 (UTC)

@Baratiiman: Your account in Wikidata is not (and never has been) blocked. What happens when you try to edit there while logged in? What happens when you try to edit while logged out? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:19, 17 September 2020 (UTC)

One issue with section editing (bug?)

Hello all! I noticed that section editing keeps track of the section with (let's call it) "section count" and no other information. So, when any section above the one you are editing gets removed (by an archiving bot, for example), instead of generating an edit conflict, the section you publish just replaces the section that has just received the same "section count" after the page updated (exhibit). Obviously I'll have to be careful about hitting publish around 05:40 UTC from now on, but it seems like something that ought to be fixed too (assuming my conclusions are correct). Would anyone please investigate and see if this is something for the phabricator? Or maybe warn about this somewhere prominently at WP:SECTIONEDIT? If it's not just me, or just the Teahouse, entire discussions may have been lost to talk page history in multiple other pages. TIA, Usedtobecool ☎️ 15:37, 17 September 2020 (UTC)

This can't really be fixed. The page will warn you if it happens to be a section which 'no longer exists' (in the sense that there are fewer sections than the section number in the URL string), which is about the best it can do. --Izno (talk) 16:27, 17 September 2020 (UTC)

Undoing list-defined references.

I've probably asked this before (and gotten an answer) but can't find it in the archives. When I originally wrote Edward G. Faile, I used list-defined references. That was before I started using Visual Editor. VE can't deal with that style, so I need to convert it to a form VE groks. What's the best way to do this? -- RoySmith (talk) 17:30, 17 September 2020 (UTC)

Ah, I found it. Help talk:Footnotes/Archive 2#List-defined references vs. Visual Editor?. Once again, google search wins the day when wiki-search fails. -- RoySmith (talk) 18:05, 17 September 2020 (UTC)

Citations button error revisited

The Citations button (edit page) is still giving me a "request failed" message. I first reported this error on Sept 8 and the final response was that it was because of load. However, I have yet to see this work regardless of time of day. Can the error message be changed so it's more accurate as to what the actual issue is? I remain unconvinced it is because of load. RedWolf (talk) 17:51, 16 September 2020 (UTC)

I've been testing it and load is definitely a factor. If one selects "thorough mode" when activating it on a large category or massive set of articles via #UCB_webform (I'm looking at you, @AManWithNoPlan:} User:Citation bot will slow down, and it will become difficult to activate Citation bot directly, by clicking the Citations button, or by clicking Expand citations on the sidebar. Abductive (reasoning) 05:35, 17 September 2020 (UTC)
Yeah, I have been testing it under load. I have made some substantial improvements and found a couple memory leaks. AManWithNoPlan (talk) 12:01, 17 September 2020 (UTC)
@AManWithNoPlan:}, is there a way to prioritize editors who are activating Citation bot from the article button/expand citations link? This is where all the frustration is happening. Abductive (reasoning) 20:03, 17 September 2020 (UTC)
I have made some changes to help with this. Long-term perhaps having two deployments might work better. AManWithNoPlan (talk) 20:05, 17 September 2020 (UTC)

Merge two talk pages with archives

We're considering merging WT:ITNR into WT:ITN but the former has been around a long time and has it's own archive. Any suggestions on how we could list the archives from WT:ITNR on WT:ITN without breaking the archive bot? --LaserLegs (talk) 19:32, 16 September 2020 (UTC)

LaserLegs, WT:ITN uses template {{archives}} to list its own archives. You could add another {{archives}} box, which would list old archives of WT:ITNR by using its parameter "root". E.g. {{archives|auto=short|search=yes|root=Wikipedia talk:In the news/Recurring items}}. This code generates the box on the right. —⁠andrybak (talk) 14:34, 17 September 2020 (UTC)
Thanks Andrybak. That worked well — Martin (MSGJ · talk) 20:16, 17 September 2020 (UTC)

"Cite" on the code editor toolbar

refToolbar's "Cite" menu is appearing on the code editor toolbar when editing a Lua module, CSS, JavaScript or JSON. At first I suspected it was interfering with a script I'd installed, but it appears even when logged out (it seems to depend on namespace though, as it doesn't appear when editing a user CSS/JS).

I don't remember this being the case (right?) and I hope this will be fixed soon, since obviously there's no use in that menu when editing a module, CSS, etc. and clicking on it does nothing (unless you switch to the plain textbox). Nardog (talk) 08:07, 18 September 2020 (UTC)

Sandbox right-clicking copy-paste issues

I have a problem. Recently I edited the Silent Scope (video game) article and I noticed something funny: while I'm at the edit page sandbox, when I try to copy an original link under the archived link on the Wayback Machine, when I right-click, the picture freezes for a few seconds up to a minute before the right-click box appears. Also, when I try to cut/copy/paste a copied link onto the sandbox, the right-click has the picture freeze up for a few seconds up to a minute before the right-click box appears! The right-clicking in the Wikipedia sandbox seems to have some issues. Does anyone know why? (I'm using Google Chrome on Windows 8.1 BTW.) --Angeldeb82 (talk) 03:06, 18 September 2020 (UTC)

Some wabsites will fire up some JavaScript when you right-click, and that may soak up processor time before the expected behaviour is allowed to continue. There are various reasons for the javascript - some newspaper websites use it to modify the text that is to be copied to your clipboard, so that when you paste what you thought was going to be the article title, you get a whole bunch of other junk as well such as a the page URL, the author, the date and time, etc. --Redrose64 🌹 (talk) 09:27, 18 September 2020 (UTC)
Well, I think I know why the right-click issue freezes up for a few seconds on other sandboxes like Wikipedia and TV Tropes: it was because of the spell-check thing that's going on in Chrome. So I disabled spell-check, and now I'm able to copy-paste without the right-click issues freezing up. --Angeldeb82 (talk) 15:10, 18 September 2020 (UTC)

WP:ITNRD on the mobile app?

Crossposting based on feedback received to post on this group.

I was playing with the mobile app (on iOS) earlier, and I was able to see WP:ITN. However, clicking on the link just shows the top three news blurbs (each as a section). Does someone know why the mobile app does not carry the 'Ongoing events' nor the 'Recent Deaths' sections from the ITN panel? Is this question better asked in a different forum? Thanks. Ktin (talk) 20:53, 17 September 2020 (UTC)

@Ktin: You say Crossposting based on feedback received to post on this group. - where else was this posted? In accordance with WP:MULTI, I don't really want to answer here if answers have already been given elsewhere.
But that aside, do you see the Nominate an article link? If that is also absent, the most likely cause is that the mobile CSS is set not to display elements belonging to the itn-footer class. --Redrose64 🌹 (talk) 09:21, 18 September 2020 (UTC)
Redrose64, thanks for the note.
Here's the link to the crosspost Wikipedia_talk:In_the_news#WP:ITNRD_on_the_mobile_app?. I was referred by folks there to attempt posting here.
I was using the Wikipedia iOS app. There is no link to "nominate an article" on the iOS app. Instead we see the first blurb show up as "in the news". There is a link below that called "current events", and that takes me to a screen with the three blurbs from the WP:ITN panel, and that's about it.
Appreciate your inputs. Ktin (talk) 15:47, 18 September 2020 (UTC)

File upload down.... again

Hello, once again the file upload wizard is down. You add a pic, fill in all the boxes, press upload, it gets to 100%... and then nothing. No file. It has been doing this on and off repeatedly for the past few months. Very frustrating. WisDom-UK (talk) 18:17, 18 September 2020 (UTC)

How large is your file (bytes, not pixels)? --Redrose64 🌹 (talk) 20:46, 18 September 2020 (UTC)

New template: autocat

Hi there, I am not totally sure, this is the right place to start this conversation, but I recently developed a Lua module on Turkish Wikipedia and translated into here. {{autocat}} works with Module:Autocat. Quickly I will explain what it does:

  1. It has no parameter. So, it takes the page title, which is "2000 works", and divides into parts: the year and the suffix. In this case it is something like: "2000", "works".
  2. And then It goes through Module:Autocat/data/arts and finds the "works" object, and uses that information to create a header bar, explanations, commons category links and portal links. And it checks if the category has a main article or see also articles.
  3. It inserts the category the appropriate parent categories.
  4. It has no parameters. Just placing the template to a category page.

The plus sides:

  1. We would not need any of the templates listed in {{Creative works chronology category header templates}} or similar ones. It would be like a cleanup. It does not seem like a good cleanup but, we have hundreds of templates in Chronology category header templates. They all would be useless after this template.
  2. We would not need to create {{WorksYear}} or {{WorksDecade}} templates seperately, they all use the same data listed on a Lua page, which is just a couple of lines.
  3. Adding this template to the category pages can be done by the bots, as the same thing done on Turkish Wikipedia.

But the catch is that the data should be entered in Module:autocat/data pages. That is not a hard thing to do, if the sample data I put is examined. Just a couple of lines for each type.

So I am proposing to use this template on these chronology categories. ~ Z (m) 17:44, 18 September 2020 (UTC)

WP:MULTI strikes again, this template was mentioned at Wikipedia talk:WikiProject Categories#Template:Autocat. --Redrose64 🌹 (talk) 18:28, 18 September 2020 (UTC)
To be fair, these two discussions were started by different people, and it's quite possible Z didn't notice my discussion before starting this one. * Pppery * it has begun... 20:56, 18 September 2020 (UTC)
This would disenable at least one well-used gadget and VisualEditor and make it difficult to remove unwanted cross-cutting categorization. As well as add to the expansion limit on a given page. This would also require a more descriptive name. --Izno (talk) 18:37, 18 September 2020 (UTC)
"the catch is that the data should be entered in Module:autocat/data pages" We already have Wikidata for that. Paging @Mike Peel and RexxS: as they have done a lot of similar work, not least on on Commons. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:41, 18 September 2020 (UTC)

Using a template from a sister project

(Originally asked in the help desk, but was referred here).

Is it possible to use a template from a sister project of Wikipedia, or from another language of Wikipedia? E.g. I need something like {{wikt:link|en|go|went}} to call Wikidictionary's template {{link}} with the aforementioned parameters, and in another language's wiki {{en:xpd|CURRENTDAY}} should call {{xpd|CURRENTDAY}}. Thanks, Veritas94 (talk) 22:23, 18 September 2020 (UTC)

No, you must import the template to the wiki of interest. --Izno (talk) 23:43, 18 September 2020 (UTC)

Inexplicable tag

Does anybody know why this edit has the tag "Reverted"? It's not a revert. --Redrose64 🌹 (talk) 18:26, 18 September 2020 (UTC)

I am uncertain whether reverted refers to the edit in question reverting an earlier revision or if it refers to a future revision reverting the so-tagged edit. I am fairly certain it is the latter given this also is so-tagged. Tags can be assigned to arbitrary revisions, it's just that there are very few things that do so. --Izno (talk) 18:45, 18 September 2020 (UTC)
It means that the edit was later reverted (in this case, in the next edit). I don't recall seeing that tag until a day or two ago; maybe it's a new feature that could have been named a little better. [Edited to add: It looks like this new feature was in the works for a while and was released in August or September. See T164307 and related tasks. I sure wish the WMF developers would fix some of the pesky bugs I'm following before introducing shiny new features, but I'm just a volunteer (and a donor...).] – Jonesey95 (talk) 19:18, 18 September 2020 (UTC)
Yes, "this edit has been reverted" fits the bill - there certainly seem to be a lot of them at the moment, and yesterday was Thursday. --Redrose64 🌹 (talk) 20:17, 18 September 2020 (UTC)
As it happens these recent changes for reverted edit taggings are by volunteer/non-WMF/WMDE developers for Google Summer of Code. --Izno (talk) 21:05, 18 September 2020 (UTC)
I noticed it, too, and appreciate it. The single word seems clear to me, though shouldn't it be linked like some other tags? —[AlanM1 (talk)]— 06:07, 19 September 2020 (UTC)
It's a new feature, see mw:Manual:Reverts#Reverted_tag. Added by Ostrzyciel as part of GSoC 2020. – SD0001 (talk) 09:19, 19 September 2020 (UTC)

RedWarn and Off-Wiki WMCS hosting

Hi all,

We wanted to get some opinion regarding the way that RedWarn currently loads its script content. There had been some controversy over a previous experimental distribution method, which was originally intended to make deploying RedWarn easier, which was removed due to privacy concerns, but over the past few months with RedWarn growing larger and larger, with hundreds of users, we want to expand by completely re-writing the RedWarn tool. This is helped experimentally by the fact that we now have access to Wikimedia Cloud Services, meaning we can implement technologies on WMF hosted servers. We know that this is substantially different to what other user script/gadgets do, but we believe using more modern web technologies can yield a significantly better user experience.

Extended content

Our transition includes using TypeScript, a popular programming language made by Microsoft, which is based on JavaScript. TypeScript compiles down to JavaScript which can be used in a browser as per usual. Along with TypeScript, we'll be using Webpack to bundle up the script. This will allow RedWarn to live in a nicely-packed singular file, but still have its full capabilities. Both of these are open-source technologies which have proved their usability and trust over the years. These will only be used for developing RedWarn and end-users will only be using the compiled JavaScript generated by the TypeScript compiler. The implementation of these technologies is ongoing as we rewrite RedWarn's codebase.

Due to the new code structure, the compiled JavaScript is going to become much more complex and practically unsuitable for human reading, so we need to reconsider how to host RedWarn. Right now, we rely on a manual update of User:Ed6767/redwarn.js based on a combined version of all our JS and HTML files. This method would be fine if it weren't for:

  • a. the amount of mess-ups that can emerge from joining all the JavaScript and HTML files, not to mention how a large script can be detrimental to the performance of MediaWiki, and other user scripts.
  • b. the fact that the MediaWiki parser begins to parse critical parts of RedWarn, adding it to many hidden maintenance categories and causing other issues (such as replacing "~~~~", requiring workarounds). Despite them being implemented, MediaWiki ignores RedWarn's no-wiki tags, likely due to how large the script is.

Our ideal option is to have the userscript load the actual bundled RedWarn script from external Wikimedia provided servers (see Wikimedia Cloud Services). We already have RedWarn's in-development web application hosted on a VPS provided by the WMF a month ago. This new method involves turning the old userscript into a loader for the actual script on the WMCS. Many of RedWarn's external scripts are also already hosted on Toolforge. Here, we can also implement checksums and other security measures to ensure the off-Wiki hosted scripts are legitimate.

Failing this, we can continue the current method, which completely relies on Ed and complex MediaWiki parser workarounds, or having a bot update the code in its own user space, similar to some user-highlighting scripts, which will require the script to be moved to a different user space. With the latter method, the script will be moved to a different user page, and all commits to RedWarn's main branch will automatically update the script based on the latest build.

In addition, as an open source project, the option always stands for users to read through all our code directly.

TLDR: Given that RedWarn now has access to servers controlled by the WMF, would it be problematic to move the code off-Wiki (but still on Wikimedia servers), given that all changes are tracked on RedWarn's GitLab, and considering checksums to verify the legitimacy of the script can be implemented? If so, would it be problematic to use a bot to update the RedWarn script (within its own userspace)?

Signed: Berrely (talk · contribs), Chlod (talk · contribs), Leijurv (talk · contribs), Ed6767 (talk · contribs) and Prompt0259 (talk · contribs) 20:06, 13 September 2020 (UTC)

Unless something has drastically changed - this is still just a personal user script that some other are importing - so it is really buyer beware. If there is plan to move this to a full gadget one day, it would be good to plan ahead now. As far as updates go, there is about 0 chance we would approve a wiki-"bot" as an interface administrator just for this task, and as a personal user script the ownership and accountability for the script belongs clearly to the user that is publishing it. If that user uses some sort of script on their end to update what is their own user subpage, we're not really going to do anythign to stop it. — xaosflux Talk 21:48, 13 September 2020 (UTC)
Xaosflux's usual hostility at user scripts aside, there's no problem with doing anything described here. However, as long as there is one "lead maintainer", I don't see why you guys want to use a separate bot account's userspace to host the script. The lead maintainer can host it in their space and update it from the master branch by running a "deploy" script. But it's understandable if you want to have fully automated CI/CD process.
The nowiki tags aren't working surely because User:Ed6767/redwarn.js contains multiple such tags. Nowiki tags can't be nested. The closing nowiki tag on line 2987 closes the one started on line 22. This can likely be worked around by substituting usages of "nowiki" in the source code itself with something like "no" + "wiki" (assuming webpack doesn't join the strings while minifying). – SD0001 (talk) 08:45, 14 September 2020 (UTC)
@SD0001: I'm actually rather welcoming of user-scripts in general, like I mentioned above the user can pretty much do whatever they want and anyone that doesn't want to use it can just move along. What I am rather unwelcoming to is having to get administrators to maintain personal user scripts that become at all popular and then are abandoned by their owners. — xaosflux Talk 03:10, 15 September 2020 (UTC)
I would have thought that making RedWarn a gadget would have been on the product roadmap. Is this something that the development team has considered and rejected? Best, Barkeep49 (talk) 19:31, 14 September 2020 (UTC)
@Barkeep49, it has been a consideration, especially now I've assembled at team to maintain it. If the community wants it to be a gadget, we won't object. If it was, however, as none of us are interface admins we'd prefer to host off-wiki on WMCS with all the correct security and privacy procedures in place so we can get updates out quicker and without massive edit requests, but hosting on-wiki is possible too. Ed talk! 19:39, 14 September 2020 (UTC)
Ed6767, that makes sense. I don't know all the ins and outs of making it a gadget but I understand that it offers considerable performance advantages over being a userscript. From what xaosflux wrote above it seems like making this change would complicate making this a gadget down the road. Is that correct xaos? Are there other reasons/advantages that the RedWarn team might want to move towards making it a gadget rather than continuing as a user script? Best, Barkeep49 (talk) 19:42, 14 September 2020 (UTC) Fixing ping of xaosflux. Barkeep49 (talk) 19:47, 14 September 2020 (UTC)
@Barkeep49, being a gadget would be preferable in terms of user friendliness as installing would be as easy as checking a box in preferences, so there'd be no need to dig into code files, which many may not prefer to do. Other than that, userscripts and gadgets are very similar in a technical nature, in terms of they both add code to Wikipedia that can be run in your browser. Now we've discussed internally, going forward, we'd prefer to make RedWarn a gadget, but that will require consensus. So, while we're on topic, would it be worthwhile to open an RfC regarding making RedWarn a gadget below? Ed talk! 20:04, 14 September 2020 (UTC)
I'm the wrong person to ask that question as all I know is just how much I don't know about the topic. Best,Barkeep49 (talk) 01:52, 15 September 2020 (UTC)
As far as turning a script in to a gadget goes, that is where most of the gadgets come from. Gadgets normally get extra scrutiny and all code changes on-wiki would need to be reviewed by interface administrators - which means there may certainly be reluctance to import instructions from off-wiki that would bypass the code-review and version-fixing components. — xaosflux Talk 03:13, 15 September 2020 (UTC)
This may be off-topic, but in terms of a blacklist for user scripts, I think something like this would do well as part of m:Partial blocks and blocking in general. There should be an option in block settings that blocks REST POST API calls only while allowing editing through the regular edit form. This would effectively disable gadgets like Twinkle and programs like Huggle while allowing the user to make constructive edits elsewhere. Aasim 04:09, 15 September 2020 (UTC)
@Awesome Aasim: I fail to find what exactly about what you said is related to RedWarn's deployment options. I suggest that you make a new section and provide the relevant context there. Chlod (say hi!) 04:20, 15 September 2020 (UTC)
I think you're trying to use the user script & gadget infrastructure in a non-idiomatic way. The existing process/expertise around scripts isn't geared to handle modern frontend practices. Years and years from now, any user script or gadget will someday be edited by a interface administrator in exclusively JavaScript form, so any nontrivial build process will complicate that. (I doubt the TS compiler outputs very nice JS, and minification will of course produce write-only code. Minification will also necessitate more extra sanitization steps to get the result past the MediaWiki parser.) As an example, all Popups did for a "build process" was concatenate a bunch of pure-JS files, and modifying it was painful enough.
I would instead recommend RedWarn be converted to a standalone tool running on Toolforge, just like Snuggle. On Toolforge, you can use any frontend technologies and not bother with any bot/deploy script stuff. From using RedWarn, it looks like the only big new thing that would need adding is a custom diff viewer, which isn't that bad since diffs already come from the API in HTML form. Users could unfortunately no longer use all their other user scripts that work on diff pages, but RedWarn could just link to the normal on-wiki diff page from the custom diff view. Yes, this would also bypass the gadget process, but I'm fine with that. This isn't really a gadget anyway, if you compare it with the rest of Preferences → Gadgets. Pinging Joeytje50, the WP:JWB developer, who probably knows all about user scripts that "take over" the browser window. Enterprisey (talk!) 09:20, 19 September 2020 (UTC)

Images with transparent backgrounds?

I have a PNG of an organization's logo. It's white on a transparent background. Is there a way to include this as the "image" parameter in a {{infobox organization}} but force it to be displayed on a dark background? If it matters, it's WP:NFC. -- RoySmith (talk) 13:47, 19 September 2020 (UTC)

@RoySmith: The way a transparent PNG works is that the image embeds in itself a particular color so that every pixel with that color becomes transparent. What's the image? What software do you use to save or edit the image? Nardog (talk) 14:46, 19 September 2020 (UTC)
Nardog, The image is https://www.nysagsociety.org/static/images/logo-light.png. I know I can manipulate the image in a photo editor to turn the background any color I want. What I'm trying to figure out is if there's a way to get the infobox template to do that, without any manual image manipulation. -- RoySmith (talk) 14:51, 19 September 2020 (UTC)
I think in the past I have wrapped such logos in div tags: <div style="background-color:black;">logo.png</div> — Diannaa (talk) 14:57, 19 September 2020 (UTC)
There's a full-color white-background version of that logo here. Nardog (talk) 15:06, 19 September 2020 (UTC)
Nardog, Ah, cool, that's even better :-) -- RoySmith (talk) 15:10, 19 September 2020 (UTC)

New gadget proposal: view short descriptions in categories

Hi all. I would like to propose user:SD0001/shortdescs-in-category (source) be made a gadget. With more than 2 million articles now having local short descriptions and countless more having them on Wikidata, shortdescs have become a great way to identify articles one may be interested in. Being able to see these descriptions while browsing categories marries together these two ways of organising and classifying content (short descriptions and categories). I have found this really helpful while working through maintenance categories (eg. CAT:NN or Category:Indian poet stubs) and trying to find articles I want to work on.

This script only has about 13 users at the moment because it is very new. The existing users are largely people working on adding shortdescs. I think it will be useful to a lot more people than the ones who're using it currently, and gadgets are a lot more easily discoverable than userscripts, hence this proposal.

It is stable and works across all browsers and skins. – SD0001 (talk) 05:04, 20 September 2020 (UTC)

Support I have just installed this and like it a lot. It greatly enhances category displays by providing annotation. Thank you for writing the script. It could possibly be bundled with the shortdesc helper gadget. Thincat (talk) 16:40, 20 September 2020 (UTC)

I can't see radio buttons for diffs in Firefox

I am no longer able to see the radio buttons for diffs in the page history. I can view a diff if it's in my contributions, but the radio buttons have disappeared. My browser is Firefox 80.0.1, but it works in Chrome and Safari. Does anyone know the cause of this problem? —Naddruf (talk ~ contribs) 19:10, 20 September 2020 (UTC)

@Naddruf: I'm on Firefox 80.0.1, and I can still see radio buttons in the page history. Two things to try:
  1. Try Wikipedia's safe mode, by adding the query string parameter `safemode=1` to the URL (e.g., https://en.wikipedia.org/w/index.php?title=Wikipedia:Village_pump_(technical)&action=history&safemode=1 )
  2. Try Firefox's safe mode, by following the instructions at https://support.mozilla.org/en-US/kb/troubleshoot-firefox-issues-using-safe-mode#w_how-to-start-firefox-in-safe-mode
Jackmcbarn (talk) 19:15, 20 September 2020 (UTC)
Also try logging out to see if the appearance changes. If it does, something in your settings or js/css files may be causing the buttons to disappear. – Jonesey95 (talk) 19:37, 20 September 2020 (UTC)

Edit summary boxes

Sometimes I am getting two edit summary boxes. Anyone know what is going on? Hawkeye7 (discuss) 03:48, 21 September 2020 (UTC)

archive

there is an issue with the mobile website archives for WP:AN and WP: ANI, you cant access them unless you know the link to them. you cant search them on the search bar as well, so I am thinking that this should be added:

WP:ANI/Archive as an example, but can be used for other noticeboards. this way, it is more accessible to mobile users, or users who prefer the mobile website than the desktop website. What do you guys think? New3400 (talk) 18:24, 15 September 2020 (UTC)

This actually pertains to the same problem discussed in #Mobile site stripping page elements below. Perhaps Module:Administrators' noticeboard archives shouldn't be using the navbox class. Nardog (talk) 07:49, 18 September 2020 (UTC)

thanks for agreeing, Nardog. i feel that this can help moblie editors. New3400 (talk) 17:32, 21 September 2020 (UTC)

Tech News: 2020-39

Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.

Recent changes

  • There is a new tag for reverted edits. For example you can see it in the recent changes feed or in the article history. It is added to edits when they have been undone, rollbacked or manually reverted to an older version of the page. [18][19]

Changes later this week

  • The number of times you can do something in a period of time on wiki is limited. This could be the number of edits per minute or the number of users you email in a day. Some users are not affected by all limits because of their user rights. They could soon see the limit even if it does not affect them. [20]
  •   The new version of MediaWiki will be on test wikis and MediaWiki.org from 22 September. It will be on non-Wikipedia wikis and some Wikipedias from 23 September. It will be on all wikis from 24 September (calendar).

Tech news prepared by Tech News writers and posted by bot • Contribute • Translate • Get help • Give feedback • Subscribe or unsubscribe.

21:26, 21 September 2020 (UTC)

Requesting help

Hi,

I have created following two RfCs at

Above RfC's (neutral part) need to be correctly visible in Wikipedia:Requests for comment/Religion and philosophy and other two project pages. While other users are saying some thing I am not getting technical aspects of the same and I am looking for proactive support in sorting out the issue.

I hope and request some one help me out.

Thanks. Bookku (talk) 10:35, 21 September 2020 (UTC)

I already explained this. The RfC statement needs to abide by WP:RFCST and WP:RFCBRIEF. The RfC statement is everything from the {{rfc}} tag (exclusive) to the next valid timestamp (inclusive). In both of those RfCs, what you have in the statements is neither neutral nor brief. --Redrose64 🌹 (talk) 13:09, 21 September 2020 (UTC)
Sorry, no take personal, You explain, I no understand. I need help from some one else, who can understand what you say and do correct things without wasting time in lecturing. So let others help, I write here to seek help from others not you, because you lecture without practical helpness.
Bookku (talk) 18:19, 21 September 2020 (UTC)
@Bookku: Nardog made a fix with these three edits, which had this effect on the RfC listings. --Redrose64 🌹 (talk) 21:35, 21 September 2020 (UTC)

Troubleshooting

Wikipedia:Troubleshooting#Other problems says to come here.
Browser used: Chrome.
Edit screen: Source (I don't edit with VE).
Problem: The editing toolbar disappears after saving an edit in one article, then trying to edit another article. TW disappears after saving one edit. I have to log out and log back in to edit another article.
This started happening almost three weeks ago. Pyxis Solitary (yak). L not Q. 11:11, 12 September 2020 (UTC)

Pyxis Solitary, I have two links for you. First, try editing in mw:safemode and see whether that helps. Second, go to mw:editor and figure out which one you're using. There are lots of "source" editors. Whatamidoing (WMF) (talk) 21:08, 16 September 2020 (UTC)
@Whatamidoing (WMF): I appreciate your response, but I am not tech savvy. (1) I tested safemode in an article I just edited where the editing toolbar was gone, and with safemode the editing toolbar appeared when I opened the Source edit screen again; (2) I'm using Extension:WikiEditor (VectorEditor) and that's what's selected in my Preferences ('2010 wikitext editor'), which is set to "Remember my last editor" editing mode. As I respond to you, I have no editing toolbar. I think Beta is posing a problem with my browser. I deleted Chrome and downloaded it again -- but the problem continues. P.S. When I click on my Notifications, I don't see them. I get a bar with slanted lines moving to the right. I know you responded to my problem because I received an email. Pyxis Solitary (yak). L not Q. 03:30, 17 September 2020 (UTC)

FWIW, I've switched to using Opera browser for Wikipedia and have not experienced any problem with the editing toolbar and TW. Pyxis Solitary (yak). L not Q. 09:32, 22 September 2020 (UTC)

Adding dashes in columns

Hi, on my list User:Encyclopædius/Language learning centre/German word list. Can somebody tell me how to add a dash between words using the advanced search feature? Also how to add a # at the start of every line (for future entries)? Isn't there some sort of * thing you type in the expression box?† Encyclopædius 12:16, 19 September 2020 (UTC)

You seem to be having tab characters between the words. Also sometimes these tab characters appear at the end of the line as well. So well I ended up fixing it myself. Here's what I used:
  • \s*$ replaced with nothingness (this removes any spaces/tabs at the end of the lines)
  • \t replaced with - (this converts tabs, which are now only there b/w words, to dashes)
The "Treat search string as a regular expression" option should be checked always.
To add # at the starts, you'd do search for ^ and replace with #.
You may find this cheatsheet useful if you want to learn regex. – SD0001 (talk) 12:44, 19 September 2020 (UTC)

Thanks, appreciate it! † Encyclopædius 13:12, 19 September 2020 (UTC)

One last thing, is there an alphabetical order option? I've been using an external site but ideally I want letters with accents to be classified as a regular letter, ü for example as u. No worries if not.† Encyclopædius 13:22, 19 September 2020 (UTC)

A better # insertion regex might be
Search for: ([\r\n])([^#\{=\r\n]) – find a newline followed by anything but #, {, =, or another newline
Replace with: $1#$2 – the found newline followed by a new # followed by the character that wasn't in the search exclusion set
For hyphen replacement:
Search for: [\t ]+\-[\t ]+ – find a hyphen between one or more whitespace characters; done this way so that the newline pair (\n\r) aren't included in the match
Replace with: –  – <space><ndash><space>
No alphabetizing option. You could write a lua module that would do the sorting, insert #, and replace hyphen with ndash all in on go with the data for the German word list kept in a separate data module – you would edit the data module to add new terms and the list would magically appear at User:Encyclopædius/Language learning centre/German word list properly ordered and separated.
—Trappist the monk (talk) 13:34, 19 September 2020 (UTC)

Thanks Trappist. † Encyclopædius 14:43, 19 September 2020 (UTC)

Just added some to User:Encyclopædius/Language learning centre/Dutch word list, I managed to add the # and the dash between words but the command ended up also adding a dash at the end of the lines. How do I remove those dashes at the end and alter the command so it only adds a dash between the words and not on the end too? \t added two dashes, I only want one between words.† Encyclopædius 10:07, 20 September 2020 (UTC)

Where the line ends with a tab character find: \t matches and so replaces it with the hyphen. Find: [\t ]+([\r\n]+) Replace: $1 will remove tabs and spaces at the end of lines. Do this first.
—Trappist the monk (talk) 13:01, 20 September 2020 (UTC)

Didn't work unfortunately.† Encyclopædius 16:52, 20 September 2020 (UTC)

What do you mean didn't work? I just tried it and it removed trailing tab and space characters from the ends of lines where they exist just as I said it would. When creating a new list, removing trailing tabs and spaces must be done before you insert dashes or you end up with trailing dashes. If you were expecting the regex that I suggested to also remove the trailing hyphen characters, I never said it would. But, it can be adapted to do that: [\t \-]+([\r\n]+). But, that version will also remove dashes that are 'between' a Dutch word and its non-existent English counterpart.
—Trappist the monk (talk) 13:29, 21 September 2020 (UTC)

I had an error occur because of a gap, fixed it now, so works! Thankyou Trappist!!† Encyclopædius 12:03, 22 September 2020 (UTC)

Translatable modules proposal

Hi,

There's a new proposal to localize Lua modules in a more modern, safe, and convenient manner: mw:Translatable modules. In the foreseeable future it will only affect multilingual sites, such as Wikidata, Commons, Meta, and mediawiki.org, but at a later time it may also be deployed on Wikipedias and other projects.

It will be great if experienced module developers could take a look at the project page, mw:Translatable modules, and its subpages, especially mw:Translatable modules/Proposed solutions.

Thanks! --Amir E. Aharoni (WMF) (talk) 14:31, 22 September 2020 (UTC)

Finding a new maintainer for OneClickArchiver

It has been 5 years since OneClickArchiver had a permanent maintainer since the previous maintainer was banned. Since OneClickArchiver is still a widely used tool, somebody needs to take up the post as maintainer for OneClickArchiver. Goose(Talk!) —Preceding undated comment added 03:37, 22 September 2020 (UTC)

Agreed. HeartGlow (talk) 03:39, 22 September 2020 (UTC)
Wikipedia:One click archiving is displayed prominently on the tool page, what's wrong with User:Evad37/OneClickArchiver or User:Σ/Testing facility/Archiver? ~ Amory (u • t • c) 10:09, 22 September 2020 (UTC)
  • A great improvement to one-click archiver would be...a multi-click archiver? So people's watchlist don't get flooded by someone clicking 50 threads one-at-a-time... –xenotalk 11:29, 22 September 2020‎ (UTC)
    xeno, Sigma's linked above does so. It just doesn't respect bot settings like I'd prefer, so you need to enter the preferred archive page name manually. Yes, I've asked for support and was rejected. --Izno (talk) 14:40, 22 September 2020 (UTC)

Invisible Pictures

Forgive me if this isn't the place to ask. I've noticed this several times. On pages like this one [21] if you click on a picture then follow the arrows to the right you will come to this picture [22] It's a nice enough picture but you can't see it when viewing the page and I didn't find it in the page edit. It has nothing to do with La Porte - it's from Kemah, a town about 10 miles south of La Porte. Why is it there and how do I find it on the page? Wiki name (talk) 13:34, 22 September 2020 (UTC)

It's in Template:Galveston Bay Area at the bottom. Nardog (talk) 13:37, 22 September 2020 (UTC)
Thanks. Who defines which towns are in the Bay Area. Webster is about 7 miles from Galveston bay. Wiki name (talk) 13:57, 22 September 2020 (UTC)
Someone should probably add the 'noviewer' class to the navbox images. —TheDJ (talk • contribs) 14:27, 22 September 2020 (UTC)
Here's a fix for {{Galveston Bay Area}}. However, this doesn't take care of plain wikitext images, like Texas flag in {{Houston-Sugar Land-Baytown MSA}} (example from the same page, La Porte, Texas). Should the whole navbox have CSS class "noviewer" to remove all decorative images from Media Viewer? —⁠andrybak (talk) 14:45, 22 September 2020 (UTC)
That seems reasonable to me in the general case, but I would guess there are some misuses of navbox out there which would cause an image to be hidden from the viewer. (I don't think that's worth spending time on.) --Izno (talk) 15:17, 22 September 2020 (UTC)
+1 for this fix. I've certainly been frustrated by images in navboxes, mboxes, flagicons, etc. showing up in the Media Viewer. If there are "misuses", then they're what need to be fixed on sight. Nardog (talk) 16:42, 22 September 2020 (UTC)
I've created an edit request: Template talk:Navbox § Template-protected edit request on 22 September 2020. —⁠andrybak (talk) 16:45, 22 September 2020 (UTC)
And done (with no prejudice to reversion in case of a sizeable opposition or better solution). Nardog (talk) 17:09, 22 September 2020 (UTC)

Collapsible wikitable not working

The collapsible wikitables on the article GOMS#CMN-GOMS (versioned link) don't seem to be working (i.e. not closable), neither on Firefox (79.0) nor on Chromium (84.0.4147.89). Is there anything that can be done about it? --Jaquento (talk) 12:53, 22 September 2020 (UTC)

It's because each table has only one row. You need at least two rows, because the top row cannot be collapsed otherwise the "[show]" link would have nowhere to be visible in. --Redrose64 🌹 (talk) 13:09, 22 September 2020 (UTC)
I see! Then it might be best to remove the "collapsible" class inside the wikitext. Thanks! --Jaquento (talk) 13:40, 22 September 2020 (UTC)
No, it'd be best to mark up that content correctly, as it is in no way a table. It looks like some sort of <code> to me, but try some flavor of Template:Syntaxhighlight, for example. — JohnFromPinckney (talk) 19:01, 22 September 2020 (UTC)

Module talk with strange errors

Tracked in Phabricator
Task T263605
Resolved

Can someone take a look at Module talk:Sandbox/Zyxw/test and see why it's generating "internal error" when one tries to even look at it? It's currently in Category:Pages using Timeline which is being renamed to Category:Pages using the EasyTimeline extension and it's not clear if this one is a cache delay or needs adjustment to be moved over. Timrollpickering (talk) 20:14, 21 September 2020 (UTC)

This line is the culprit:
self:preprocess_equals('<templatedata />', '', {nowiki=1})
It looks like <templatedata /> cannot be preprocessed. This reproduces the error:
local p = {}

function p.run_tests(frame)
	return frame:preprocess('<templatedata />')
end

return p
Nardog (talk) 21:11, 21 September 2020 (UTC)
I commented out the line causing the error and now the page is no longer in that category. -- Zyxw (talk) 22:57, 21 September 2020 (UTC)
I filed phab:T263605 for the general bug, even though the specific case is resolved. * Pppery * it has begun... 00:58, 23 September 2020 (UTC)

How does notifications ignore work?

In the Notification section of my Preferences, I've added The Only Way Is Essex (series 26) to the "Muted pages for page link notifications" (I created it as a redirect, and now that it's an article, I was getting notifications about it). But I just received a notification that the page was tagged with link rot- this shouldn't have happened. Does the muted pages take some time to work? I only added this page to the muted pages about 4 hours beforehand. Or is it a bug? Joseph2302 (talk) 12:33, 23 September 2020 (UTC)

Edit conflicts with yourself?

Do edit conflicts only get detected between different users? I've been trying to force edit conflicts (so I could test some aspects of how they behave) and was surprised to discover I couldn't generate one by editing User:RoySmith/sandbox logged in as myself in two different windows. If I opened an incognito window and logged in as RoySmith-testing, I could then easily generate conflicts. -- RoySmith (talk) 23:49, 22 September 2020 (UTC)

In theory, it should be impossible to conflict with yourself. The later edit will just save, with no attempt at a three-way merge. In practice, it can sometimes happen, as the above section shows. My guess is it only happens when the edits are submitted a fraction of second apart, but I don't really know. Suffusion of Yellow (talk) 23:51, 22 September 2020 (UTC)
On second thought, it looks like a self-conflict can happen during a section edit. From EditPage.php:
             // An edit conflict is detected if the current revision is different from the
             // revision that was current when editing was initiated on the client.
             // This is checked based on the timestamp and revision ID.
             // TODO: the timestamp based check can probably go away now.
             if ( ( $this->edittime !== null && $this->edittime != $timestamp )
                 || ( $this->editRevId !== null && $this->editRevId != $latest )
             ) {
                 $this->isConflict = true;
                 if ( $this->section == 'new' ) {
                     if ( $this->page->getUserText() == $user->getName() &&
                         $this->page->getComment() == $this->newSectionSummary()
                     ) {
                         // Probably a duplicate submission of a new comment.
                         // This can happen when CDN resends a request after
                         // a timeout but the first one actually went through.
                         wfDebug( __METHOD__
                             . ": duplicate new section submission; trigger edit conflict!" );
                     } else {
                         // New comment; suppress conflict.
                         $this->isConflict = false;
                         wfDebug( __METHOD__ . ": conflict suppressed; new section" );
                     }
                 } elseif ( $this->section == ''
                     && $this->edittime
                     && $this->revisionStore->userWasLastToEdit(
                         wfGetDB( DB_MASTER ),
                         $this->mTitle->getArticleID(),
                         $user->getId(),
                         $this->edittime
                     )
                 ) {
                     # Suppress edit conflict with self, except for section edits where merging is required.
                     wfDebug( __METHOD__ . ": Suppressing edit conflict, same user." );
                     $this->isConflict = false;
                 }
             }
Suffusion of Yellow (talk) 00:01, 23 September 2020 (UTC)
Suffusion of Yellow, Interesting, thanks. But, to take a step back, what I'm really trying to figure out is what happened with this edit. I edit-conflicted with Graywalls, thought I had properly resolved it using the Visual Editor's EC-resolution tool, but what ended up happening was deleting the white space between paragraphs, which ran them together. See this perma-linked version.
I've been trying to figure out what happened, but I've been unable to reproduce that behavior. I can now get edit conflicts (by using two accounts), but I still can't reproduce the whitespace removal. -- RoySmith (talk) 00:17, 23 September 2020 (UTC)
My situation (above) is very simple. I edit and I get an edit conflict. I see no other adjacent activity from anyone else in history. I will try to determine whether this only occurs on section edits. ~Kvng (talk) 15:08, 23 September 2020 (UTC)
I'm getting edit conflicts with self on non-section (whole article) edits. ~Kvng (talk) 16:02, 23 September 2020 (UTC)

Is there a script that highlights or corrects redirects?

Hi all, I've been editing navboxes recently and quite a few terms link to redirect pages. I was wondering if there is a script that could highlight these or fix them for me? This would be particularly useful because often due to moves and mergers terms used may be out of date, or worse there may be several terms linking to the same page, and I'd like an easy way to work out which ones I need to focus on. Does anyone know if there is such a script already? Cheers --Tom (LT) (talk) 04:21, 23 September 2020 (UTC)

For highlighting, see User:Anomie/linkclassifier. Headbomb {t · c · p · b} 04:43, 23 September 2020 (UTC)
Helpful! Thanks. --Tom (LT) (talk) 06:40, 23 September 2020 (UTC)
@Tom (LT): Something a little less heavy for highlighting/require less setup would be similar to what I have in User:Izno/common.css (.iznoredirects is for temporary addition to a long table or page that could use some redirect resolution, so you can ignore that line if you prefer). --Izno (talk) 17:23, 23 September 2020 (UTC)

Randomising names in a list

G'day all, I have a situation where there is disagreement about what order some names should go in the infobox of an article about a war. Reliable sources are not very helpful in determining an order of importance and of course there are different factors to consider where one leader may be more important in one respect, and another leader more important in a different respect. My question is, is there a way of formatting a plain list within a field in an infobox that would present the names in a random order each time it was viewed? Thanks, Peacemaker67 (click to talk to me) 03:08, 24 September 2020 (UTC)

Sure can!
{{#invoke:random|bulleted_list|egg|sausage|spam}}
gives
  • sausage
  • spam
  • egg
See Module:Random for more tricks. Hawkeye7 (discuss) 03:31, 24 September 2020 (UTC)
Thanks Hawkeye, I shouldn't be surprised that you would know how to do it. Peacemaker67 (click to talk to me) 03:46, 24 September 2020 (UTC)

Interesting idea, and it begs the question about unconscious ordering bias in lists. There probably is an attention bias towards the first and last entries in a list. It is a testable hypothesis. And feuds over pecking order show there is a supposed bias editors are trying to leverage. A is better than C better than F .. unless F is the anti-hero. And so on, it's ingrained in mental maps. Could see it being used for things like award shortlists. -- GreenC 03:38, 24 September 2020 (UTC)

Thanks, yes it is an issue in some situations, GreenC. In this case there is considerable systemic bias involved IMHO, as well as feuding sides... Peacemaker67 (click to talk to me) 03:46, 24 September 2020 (UTC)

The names could be randomized, as is done with the arbitration committee elections, for example. But I think it is a bad idea to have randomly changing lists in articles. I think it would be better to find some neutral order, like alphabetical. isaacl (talk) 03:40, 24 September 2020 (UTC)

The problem with alphabetical is that reverse alphabetical is often the response, and both benefit one side or another in a dispute. Peacemaker67 (click to talk to me) 03:46, 24 September 2020 (UTC)
The guidance from the manual of style is the "most basic form of organization is alphabetical or numerical". There isn't really a good argument for reverse alphabetical in an encyclopedia article. isaacl (talk) 04:06, 24 September 2020 (UTC)
OK, I see you're talking about the leaders of World War II... The country names really should be listed, not just flags, and maybe some numeric criterion (as I see has been suggested on the talk page) based on the countries could be agreed upon. isaacl (talk) 04:18, 24 September 2020 (UTC)
Correct, but the issue is that consensus on a metric doesn't seem likely. All sorts of factors are in play. Peacemaker67 (click to talk to me) 04:31, 24 September 2020 (UTC)
Status quo can stay in place, then, until enough editors agree on something. These are really short lists; it doesn't make much practical difference. isaacl (talk) 04:34, 24 September 2020 (UTC)
It's bad for the website caching to need to deal with a random order. Please generally avoid that in live articles. --Izno (talk) 04:12, 24 September 2020 (UTC)
What practical effect is there? Peacemaker67 (click to talk to me) 04:17, 24 September 2020 (UTC)
Because of caching, the order remains the same until the page is edited or purged. Caching takes precedence over randomisation, so randomisation in articles shouldn't have an adverse effect on Wikipedia's performance. — Mr. Stradivarius ♪ talk ♪ 12:16, 24 September 2020 (UTC)
<Queue people making small edits until the cache contains their preferred version.>
I think randomizing lists in a live article is not a good idea. Two people accessing the same revision of an article should see the same content, in the same order. –xenotalk 12:20, 24 September 2020 (UTC)
I agree. Don't see anything wrong with alphabetical, this strikes me as a solution in search of a problem. Nardog (talk) 14:05, 24 September 2020 (UTC)
For articles, don't think this should be attempted - the presentation order of content is a purely editorial matter and this sort of structure will make it harder for editors to manage the content - especially when using visual editor and expecting a WYSIWYG result. — xaosflux Talk 14:37, 24 September 2020 (UTC)
Random would seem to provoke the reader to question the meaning of the order, and why it would change. Then there's the performance issue, too. Alpha is reasonable. Another option in a list of belligerents is by date of engagement. —[AlanM1 (talk)]— 19:38, 24 September 2020 (UTC)
  • Agreed that alphabetical should be sufficient. Feel free to take the consensus from here to the page; unless there's some important context we've missed, it overrides the more local consensus per WP:CONLEVEL. I'd add that it's confusing for readers to see one order and then find a different order next time they visit the page. {{u|Sdkb}} talk 20:19, 24 September 2020 (UTC)

Page views by Georgraphy

Hello all:

I was hoping to do an analysis, and was looking for geographical page view data -- is there some place where I can get it?

E.g. This link [23] or this one [24] has the page view data for an article / page. Is there some place where I can get the next level breakdown by geography as well?

Thanks, Ktin (talk) 00:13, 20 September 2020 (UTC)

@Ktin: No, this information is not publicly available for privacy reasons. Sorry. — MusikAnimal talk 20:59, 24 September 2020 (UTC)
MusikAnimal, makes sense. Thanks. Would this be available for pages like the main_page? Ktin (talk) 21:16, 24 September 2020 (UTC)
@Ktin: Country data is available for projects as a whole, see https://stats.wikimedia.org/#/en.wikipedia.org/reading/page-views-by-country/normal%7Cmap%7Clast-month%7C~total%7Cmonthly. I think offering country-level data for any given page is unlikely to happen. However, there is talk of creating a list of the most popular pages per country (phab:T207171), and those lists would certainly include the Main Page. I imagine the end result would be something like toolforge:topviews but with a way to filter by one or more countries. — MusikAnimal talk 22:05, 24 September 2020 (UTC)

Changes lost on accidentally navigating away from edit window

Every time I accidentally navigate out of the edit mode (such by clicking on a link in the preview), and go back, I find all my changes are lost and I've to start from scratch, which is extremely frustrating. I'm pretty sure this wasn't the case some time back – the edits used to get recovered. I'm wondering if it's something wrong with wikipedia or at my end? I'm using Chrome on Windows, if that matters. – SD0001 (talk) 07:01, 25 September 2020 (UTC)

There's an editing preference for that: "Warn me when I leave an edit page with unsaved changes". Also, on some browsers, you can go back to earlier pages including showing the wikitext in the edit box that you never saved. Johnuniq (talk) 07:44, 25 September 2020 (UTC)

Birthday boy

In the infobox for Joe Wicks (coach) there is a line of coding which reads

{{Birth date and age|1985|09|21|df=y}}

This generates his age as "34" but I would have thought it should be "35". After playing round with the coding I managed to get it to generate "35" but it's gone back to "34". If I enter his birthday as 20 September it gives age 35, so why does it fail to recognise his 35th birthday? The current age can also be generated using

{{birth date and age|1985|09|21}}

This generates age 35 in preview. According to [25] Joe was born in 1986 so the article age is correct, and the dab page also says "1986". In 2016 the article had 1985 in the lead and 1986 in the body. In 2018 the lead still said 1985 and this was the only reference. On 22 August 2019 someone changed "Category:1986 births" to "Category:1985 births". On 11 March 2020 the birth year was corrected to 1986. This year was shown when the infobox was added. On 23 June someone changed the birth year to 1985 in both places. I haven't edited the errors out because the issue of how an infobox can give an age of 34 when "21 September 1985" is coded remains unresolved. 92.19.171.232 (talk) 11:59, 21 September 2020 (UTC)

The Companies House data gives 1985 so I would be inclined to leave things as they are (provided we can get the article to show his age as 35 rather than 34). 92.19.171.232 (talk) 12:08, 21 September 2020 (UTC)

I just purged the page and now it says 35. Nardog (talk) 12:14, 21 September 2020 (UTC)
Thanks. 92.19.171.232 (talk) 12:19, 21 September 2020 (UTC)
Anne Robinson and Serena Williams could also do with being purged. Is this a widespread phenomenon? 79.73.15.145 (talk) 15:56, 26 September 2020 (UTC)
Anne's age is now correct. 79.73.15.145 (talk) 15:57, 26 September 2020 (UTC)

Using GitLab for code review

As was recently mentioned in Tech News, the foundation is currently holding a consultation about moving development from Gerrit, the current code and patch review system, to GitLab (for reasons, other systems are not being considered). I'm in the working group steering this consultation and we are also very interested in the opinions of those outside the core development team. Bot writers, gadget developers etc. Do you have concerns or enthusiasm about gitlab ? Do you think you might contribute more or less or even if it might be easier for you to be informed with Gitlab instead of gerrit ? What do you think about the gerrit patch review system vs the gitlab pull request system ? Please take a minute to think about it and maybe leave some comments at mediawiki.org or if you really prefer to do so, leave them here and I will later summarise them there. —TheDJ (talk • contribs) 07:44, 21 September 2020 (UTC)

GitLab is a million times better than Gerrit. ProcrastinatingReader (talk) 13:20, 21 September 2020 (UTC)

Speaking of which, is it just me who find Phabricator a pain in the butt?

  • You have to learn a markup language that's neither MediaWiki's nor Markdown and seemingly has no application outside it;
  • the way to mark a task as a bug report is somehow not through adding a tag and hidden in the task creation form itself, not in the dropdown menu at the top (which I just learned can be configured), and it can't be done or undone retroactively (or can it be? if so, that proves my point);
  • the list of your subscriptions is not accessible from the profile page or menu but only from "Tasks & Bugs", and you have to make new queries if you ever want to change the grouping or sorting;
  • (un)subscription is public and gratuitously floods tasks' timelines;
  • mentioning someone adds them as a subscriber, so if they're not interested they have to unsubscribe, which again gratuitously populates timelines and could potentially discourage users from pinging others, not a good design for something whose whole purpose is collaboration;
  • no one knows what the "tokens" are for.

It looks like it has a rich set of features, which is probably why it was chosen in the first place, but the UI is so impenetrable it repels you. The consultation page says GitLab's issue tracker wouldn't be used, but then Phabricator could also use some improvement if we want to lower the bar to contribution IMHO. Nardog (talk) 19:06, 26 September 2020 (UTC)

All of those critiques of Phab seem like paper cuts. You should use Phab to report them... of course ;). --Izno (talk) 19:34, 26 September 2020 (UTC)
Not just you, phab also sucks. Trust enterprise to choose obscure, bad solutions to things. ProcrastinatingReader (talk) 20:03, 26 September 2020 (UTC)

Copy-and-paste translation problem in editable content of non-English language

Hello VP, yesterday I created an article "List of Oteckovia episodes". I wanna expand "Episode" section, with bring "Part" section information from its Slovak language for TV series Oteckovia. Unfortunately, my Google Translation couldn't translate it to English language as I entered editing session. I desired to copy-and-paste the Slovak language when words completely translated to English, but I can't do this. I also raised this at help desk for their help. I acknowledged that this community stands for technical issues. In short, can you do the following by yourself.

1. Copy the Part section of listed episodes in List of Oteckovia episodes to List of Oteckovia episodes
2. Remember that this copy-and-paste method is complied with translation to English.

Thank you The Supermind (talk) —Preceding undated comment added 22:40, 26 September 2020 (UTC)

Infoboxes not using Template:Infobox mapframe?

Just found out about {{Infobox mapframe}}, which seems to be used in a lot of infoboxes. It seems much better than the old way, where coordinates were displayed on maps that could only be set from specified modules with coordinates and scale of the map baked in. But it isn't used, for example, in {{Infobox island}}. What's up with that? What would need to happen in order for the islands infobox to be update to use a mapframe (instead of the old one)? Is there a reason why we've chosen to keep the old one, or is it just that nobody's gotten around to all the infoboxes yet? { } 06:47, 25 September 2020 (UTC)

No-one has gotten to implementing the recent RFC on mapframes. --Izno (talk) 13:00, 25 September 2020 (UTC)
I've been adding them to some slowly, like station and settlement. Due to Q2 in the RfC, some talk discussion needs to be had on the location I think, which slows the process down a bit. A consensus to default to pushpin location, if exists, might've been handy. ProcrastinatingReader (talk) 20:07, 27 September 2020 (UTC)

Visual editor sometimes completely stuck on loading

This seems to have just started happening a few weeks ago. Are there things in my settings or preferences I should reset to default in case it is something I have changed such as adding scripts? Chidgk1 (talk) 07:28, 26 September 2020 (UTC)

@Chidgk1: I would suggest trying the steps mentioned at mw:Help:Locating broken scripts. If it still doesn't work, you'll want to file a bug on Phabricator. Kaldari (talk) 23:10, 27 September 2020 (UTC)

Questions regarding page views / analytics

Hello all:

Had a few questions regarding page views / analtyics, and was wondering what was the best place to ask those questions. I know, I had posted a question re: geographic slices of traffic, and MusikAnimal had helped answer, that we didn't have that ability.

Follow-up questions:

1. Is there a way to get referral information for a page. Specifically % page views by referring page. E.g. %page views coming in from Main_Page, %page views from Article A etc.

2. Also, I was slicing some comparative analytics and this chart caught my attention. link to chart. Do we know what happened at the end of April that caused that dip? If this was a change in measurement, did that cascade to the article pages as well?

Greatly appreciate pointers to a different group, if that is a better place to ask these questions.

Thanks, Ktin (talk) 01:05, 28 September 2020 (UTC)

  1. Nope. There might be a Phab task somewhere for it.
  2. Yes. Automated was created. The views associated were accordingly recategorized.
--Izno (talk) 01:13, 28 September 2020 (UTC)

Self edit conflict

I'm having frequent edit conflicts with myself. This has plagued me previously and then mysteriously resolved itself. It is back. I asked for help previously and found that I'm not the only one with this issue and on one has a cure or workaround. My only recourse is to resolve the edit conflict by copying and pasting the entire article text. One previous suggestion was to reset my account preferences to defaults. I've done this and the problem persists. ~Kvng (talk) 14:39, 22 September 2020 (UTC)

Please check what happens by using the following procedure to publish/save edits after previewing them: in the edit summary box, write a message then press Enter. That will save the edit (assuming you haven't done anything to disable the default behavior). The aim is to determine whether that gives a different result from using a mouse to click Publish changes. Johnuniq (talk) 23:35, 22 September 2020 (UTC)
@Kvng: (1) Are you using WP:wikEd (would be enabled under Special:Preferences#mw-prefsection-gadgets)? (2) If not, are you double-clicking the "publish" button? (3) If not, do you perhaps have an failing pointing device that's sending spurious double clicks? Suffusion of Yellow (talk) 23:40, 22 September 2020 (UTC)
I have wikEd enabled in preferences but I leave it disabled except temporarily when I need one of its features. I have definitely gotten these edit conflicts with wikEd disabled. I am definitely not double-clicking the "publish" button and if I had a pointing device issue I'm sure I would have noticed it on other applications. To publish edits I use Enter in the summary box or the Shift-Alt-S shortcut. I may occasionally use the mouse. I will see if I can make any connections between save method and edit conflicts. ~Kvng (talk) 16:00, 23 September 2020 (UTC)
I did not get any self edit conflicts yesterday or this morning. This is the same pattern as previously. The problem just mysteriously goes away at some point. ~Kvng (talk) 16:37, 25 September 2020 (UTC)
Now two full days and 50 edits with no issues. This has been fixed through no action on my part. I sure would like to know what it is because, based on past experience, I'm confident it will be back. The only theory I have is that it is associated with datacenter failovers. I beleive these edit conflicts happen after Wikipedia has been put into read-only mode for this sort of maintenance and testing. ~Kvng (talk) 19:04, 27 September 2020 (UTC)
@Kvng: I suspect some of the conflicts were caused by wikEd. WikEd, for me, really does try to submit every edit twice, but usually (not always) the browser blocks one of the submits. See Wikipedia:Village_pump (technical)/Archive 183#Nonexistent edit conflicts. I have no idea about the others; perhaps another script is double-submitting? I also don't know why the problem comes and goes; it could just be the clustering illusion, or it could have something to do with the type of edits you are making at the time. @Cacycle: Any possibility of looking at the patches in the archived thread? Suffusion of Yellow (talk) 21:55, 27 September 2020 (UTC)
Who knows, but intermittent issues are another sign of failing hardware. I was a skeptic on that point but became a believer when I gave that advice (try another mouse) to someone here many months ago. A week after that, I started getting self-conflicts with no other signs of mouse trouble. I have spares and changing the mouse eliminated the problem. Johnuniq (talk) 23:31, 27 September 2020 (UTC)
What I get from the archived discussion is
  1. I'm not crazy
  2. My mouse is probably fine
  3. Wiked is suspect
What I don't get is
  1. Do I need to disable Wiked in preferences or if just disabling it in the edit interface is sufficient? I suspect, based on my experience that the former is required. Sad.
  2. Why does this come and go? ~Kvng (talk) 01:28, 28 September 2020 (UTC)

Distinguishing between variables defined and variables defined and set to something

I'd like to design a template that behaves in the following way:

How do I code this? I'm not sure how conditional expressions like #if handle the difference between just defining "header" (as in the second example) and defining it and setting it equal to something (as in the last example). {{u|Sdkb}} talk 23:20, 27 September 2020 (UTC)

There is no such thing as a parameter defined without being set. {{subst:the_template|header}} is equivalent to {{subst:the_template|1=header}}. * Pppery * it has begun... 23:27, 27 September 2020 (UTC)
@Pppery: I'm not sure I have the terminology quite right with "defined" and "set". But what I'm trying to do is the above. Are you saying that that is impossible? I feel like I do recall some templates that have behaved differently between {{subst:sample_template|1=foobar}} and {{subst:sample_template|foobar}}. {{u|Sdkb}} talk 23:43, 27 September 2020 (UTC)
No, it's not impossible. My point was more that what you are doing is inventing a non-standard parameter convention, while phrasing it as if it were standard. What you are trying to do can be accomplished by {{{header|{{#ifeq:{{{1}}}|header|DEFAULT_HEADER|NO_HEADER}}}}}. However, I think that syntax is confusing, and it would be cleaner to do something like {{yesno|{{{header|}}}|no=NO_HEADER|yes=DEFAULT_HEADER|def={{{header}}}}}, which turns "|header=yes", and other yes-like values (as defined by Template:Yesno) into the default header, and turns |header=no and other no-like values (including the empty string) into no header, while leaving any unrecognized values as it. It's somewhat difficult to discuss this in the abstract, so it may help to point to a specific template you are trying to implement. * Pppery * it has begun... 00:03, 28 September 2020 (UTC)
Pppery, thanks! I'll go with the cleaner suggestion. The application is to a few different templates in the user warning space, such as {{Ping fix}}. {{u|Sdkb}} talk 04:13, 28 September 2020 (UTC)

Tables, row headers, and accessibility

Hello all,

Per WP:DTT, row and column headers are recommended for use in tables used in articles on Wikipedia. However, it's unclear whether this is a suggestion or a recommendation or a mandatory requirement. I personally think that both should be made mandatory unless there is some very good reason not to include them. This will help ensure that people whose screen reader software does not already associate the rows/columns with the first cell in such row/column are able to easier understand the content of a table. I am not proposing a mass removal of tables without headers or a mass change of tables currently used in articles either, but I do feel that it should be clarified in the MOS, fixed when it is happened upon by an editor aware of it, and be grounds for adding headers to tables already in place.

One issue with this proposal is that currently, row headers are rendered by most browsers as bold, with a different background, and with centered text. For a variety of reasons, it is often desirable to have the row header simply display as a regular cell - for which the table class of "plainrowheaders" exists. Unfortunately, however, this only removes the bolding and removes the forced centering of text - it does not change the background color to be that of a normal cell. This means that even when the table has a class of "plainrowheaders" there is still a noticeable and sometimes undesirable difference between the row header and the rest of the row. For this reason, I recommend that row header cells be given the "background: #f8f9fa;" style when the table as a whole uses the "plainrowheaders" class. This could (to my knowledge) be accomplished by changing the below section in MediaWiki:common.css. For the main reason of the appearance and the likely disagreement as to requiring row headers if there is no easy way to limit their change in appearance, the first proposal is contingent on this one.

Changes proposed, in more detail.

Current MediaWiki:common.css 

/* Normal font styling for wikitable row headers with scope="row" tag */
.wikitable.plainrowheaders th[scope=row] {
	font-weight: normal;
	/* @noflip */
	text-align: left;

Proposed MediaWiki:common.css 

/* Normal font styling for wikitable row headers with scope="row" tag */
.wikitable.plainrowheaders th[scope=row] {
	font-weight: normal;
	background: #f8f9fa;
    /* @noflip */
	text-align: left;

Current MOS on tables §Overview of basics§Row & column headers Like the caption, these help present the information in a logical structure to visitors. The headers help screen readers render header information about data cells. For example, header information is spoken prior to the cell data, or header information is provided on request.

§Overview of basics§Layout of table headers As can be seen in the example above, row headers are formatted by default as bold, centred and with a darker background. This is the common behavior across the Internet, and the default rendering in most browsers. In some circumstances it might be desirable to apply a style customized for a specific case. The class plainrowheaders will apply left-aligned and normal-weight formatting so that editors do not feel the need to override the header formatting with inline CSS declarations for each cell. Used by itself, plainrowheaders will make headers appear similar to unmodified data cells, except for the darker background.

Proposed MOS on tables §Overview of basics§Row & column headers Like the caption, these help present the information in a logical structure to visitors. The headers help screen readers render header information about data cells. For example, header information is spoken prior to the cell data, or header information is provided on request. Row and column headers should be used in all data tables present in articles.

§Overview of basics§Layout of table headers As can be seen in the example above, row headers are formatted by default as bold, centred and with a darker background. This is the common behavior across the Internet, and the default rendering in most browsers. In some circumstances, it may be preferable for row headers to be displayed as a normal cell. The class plainrowheaders will modify the table to cause row headers to appear as a normal cell - so that editors do not feel the need to override the header formatting with inline CSS declarations for each cell.

I tried looking a lot to see if there had been any reasoning for not including this in the class when it was created, but I couldn't find any. I will post at MediaWiki talk:common.css and at the talk pages of the relevant MOS pages regarding this proposal - but I'm not sure if there's more I should do for proposing such a large change. Thanks in advance for considering/discussing. -bɜ:ʳkənhɪmez (User/say hi!) 02:44, 27 September 2020 (UTC)

I support the lighter row-header background color for plainrowheaders class. But only if the data cells' background color is white as with class=mw-datatable. There needs to be contrast between header cells and data cells. [Later note: I was falling into the desire of some MOS warriors to have only way to do many things. Choice is better. People can choose to use class=mw-datatable or not. To distinguish header cells or not. It makes no difference to blind users using screen readers. They can't see the background color. Their screen readers see <th> and scope=row. But the lighter row header background color is better for the visually impaired, and for people like me. See discussion.]
Berchanhimez, please see related discussion:
Talk:COVID-19 pandemic by country and territory#Pandemic by region table. Format choice
From: Template:Color box {{color box|color|text|text color}}:
 f8f9fa - proposed background color by Berchanhimez for plainrowheaders 
Table below is from the table button in the wikitext advanced edit bar.
  • class="wikitable"
Caption text
Header text Header text Header text
Example Example Example
Example Example Example
Example Example Example
  • Added class=mw-datatable to table below.
Run cursor over rows to see rows highlight. Note that the data cells are white. I prefer that for better contrast when screen brightness is turned down. When screen brightness is turned down everything turns grayer. This added grayness (for whatever reason) can make things difficult for the visually impaired. It bothers me too. I keep screen brightness turned down to 80% almost all the time except for video viewing.
Caption text
Header text Header text Header text
Example Example Example
Example Example Example
Example Example Example
  • Added class=plainrowheaders and scope=row to table below.
Caption text
Header text Header text Header text
Example Example Example
Example Example Example
Example Example Example
  • Same table but without class=mw-datatable
Caption text
Header text Header text Header text
Example Example Example
Example Example Example
Example Example Example


  • class="wikitable mw-datatable plainrowheaders". Also added proposed row header background color (#f8f9fa).
Caption text
Header text Header text Header text
Example Example Example
Example Example Example
Example Example Example
--Timeshifter (talk) 05:02, 27 September 2020 (UTC)
  • Not sure I like the distinction between a row header and a column header, when both are headers. However, I just found out about mw-datatable which is really good. Not sure why the default isn't with this set on, and any table which is of type presentation can set it to that which will disable this. presentation is a far less common type. This can even be backwards compatible with making a new default class of wikitable2. --Gonnym (talk) 10:18, 27 September 2020 (UTC)
    @Gonnym and Timeshifter:, Thanks for the comments. I'll point out that nobody is going to be forced to use this class - just as nobody is now. You're also correct that in many cases a row header should be distinguishable from the surrounding cells. This is especially true for many of the tables on the COVID-19 pandemic by country and territory page. However, there are also times where the first cell in a row should be labeled as a header, but it does not need to be distinguishable from the other cells in the row - that's the entire reason for the class plainrowheaders to exist. I am merely proposing that we make "plain row headers" do what it suggests it should. Right now, "plainrowheaders" results in "normal text on a different background" - which is definitely not plain. The only current way for an editor to use row headers (for accessibility) and to have them look like normal cells involves each cell getting its own style="background: #f8f9fa;" - which is not only less efficient and harder for editors, but it makes the edit window a lot harder to use.
    Personally, I think it may be an issue because I have seen quite a few tables with "scoped column headers" (i.e. where they're properly constructed with scope="col" explicitly) but not the same for rows. I think a large reason for this may be due to the fact that the editor does not wish to have "row headers" highlighted over regular cells in any way, thus they try class="plainrowheaders", see that it doesn't work (i.e. it leaves the background color changed), and so they give up and don't use row headers. Again, I'm not proposing to change the default in any way - just to change this one class to actually do what it says it will do (make them plain cells). I will say I don't know that adding more table classes is going to work, and in fact I suspect it may be detrimental - I still have no clue how many table classes there actually are, but I know it's somewhere upwards of 10 and likely many more, and I have to look up anything other than wikitable mw-sortable floatright to use them right. Adding another class will just likely hardly get used and thus not do much for this issue. -bɜ:ʳkənhɪmez (User/say hi!) 10:56, 27 September 2020 (UTC)
    CSS allows us to style anything anyway we like. That does not mean we should. Accessibility isn't only for people with sight impairment. If you create a table which has the table header cell looks exactly like a table data cell, that also makes my reading and understanding of it harder, and I can see it. The fact that editors don't use scope or proper code isn't new. A lot of users don't know and also a lot don't care. Take a look a most reality television articles and you will see all kind of tables - most of which aren't accessible. Just to make my point "official", I'd oppose changing how plainrowheaders currently work. --Gonnym (talk) 11:05, 27 September 2020 (UTC)
    Gonnym, Thanks. I was also pointed to a discussion from 2010 about it so apparently I'm not the first to have this idea. Sorry for the time waste here guys, but thanks for opining regardless. -bɜ:ʳkənhɪmez (User/say hi!) 14:46, 27 September 2020 (UTC)
    I'm (still) with Gonnym; I oppose the suggested changes. Berchanhimez says, ... in many cases a row header should be distinguishable from the surrounding cells. I believe the correct statement is, "in all cases a row header should be distinguishable from the surrounding cells." — JohnFromPinckney (talk) 18:22, 27 September 2020 (UTC)

I struck out some of my first post, and added a later note. I was falling into the desire of some MOS warriors to have only way to do many things. Choice is better. People can choose to use class=mw-datatable or not. To distinguish header cells or not. Background color makes no difference to blind users using screen readers. They can't see it. Their screen readers see <th> and scope=row.

Regardless of whether class=mw-datatable is used or not, the lighter row header background color is better for the visually impaired, and for people like me. See discussion. I believe that some MOS warriors don't care about this problem for the visually impaired and for me. They would rather have conformity, and no change in "The Rules". Even if the supposed reason for these accessibility rules is to help the visually impaired. I do not have the more severe definition of visual impairment: Impairment that approaches functional blindness. I just have normal vision problems of nearsightedness, farsightedness, presbyopia, and astigmatism. And my eyes can't handle screen brightness set at 100%.

The bottom line is that the left table below is much better than the right table. For the detailed discussion about shading, bolding, blue links, contrast, and WCAG standards, see this discussion.

Caption text
Header text Header text Header text
Example Example Example
Example Example Example
Example Example Example
Caption text
Header text Header text Header text
Example Example Example
Example Example Example
Example Example Example

--Timeshifter (talk) 04:27, 28 September 2020 (UTC)

WP:GA and WP:PR toolbox template link problems

Toolbox
GA toolbox
Reviewing

Hi all, the two templates {{Peer review tools}} and {{Good article tools}} used in WP:GA and WP:PR seem to contain dead links. It seems like this is due to a problem on toolserver, maybe because the tools have moved URL? I was hoping a technical boffin around here who is in-the-know may be able to help out by correcting the links. Many thanks! --Tom (LT) (talk) 11:29, 17 September 2020 (UTC)

Which links in particular? (To wit, I've tried to remove the ones I know of that have caused issues previously.) --Izno (talk) 16:36, 17 September 2020 (UTC)
@Izno All of them. To me at least, they all redirect to 404 pages. --Tom (LT) (talk) 23:20, 18 September 2020 (UTC)
Ping. Still a problem that seems to be apparent for all peer reviews and good reviews. --Tom (LT) (talk) 04:07, 23 September 2020 (UTC)
You will need (apparently) to achieve consensus to remove the links or get in contact with the Toolserver project owner. --Izno (talk) 17:27, 23 September 2020 (UTC)
Toolserver has been down, permanently, since mid 2014. There's plenty on that matter in the archives of this page. I'm surprised it's taken six years for these links to be noticed as being dead. FWIW the "Toolserver project owner" was Wikimedia Deutschland. --Redrose64 🌹 (talk) 22:21, 23 September 2020 (UTC)
... I should have said Toolforge, the links to which do function in normal cases. --Izno (talk) 23:19, 23 September 2020 (UTC)
Great. I don't see the use in revitalising these links if noone has actually even tried to use them in the last 5 years - they clearly aren't seen as useful. I've commented out the links in both templates and, as I'm active at peer review, have also boldly added a link to Earwig's. --Tom (LT) (talk) 04:10, 24 September 2020 (UTC)
@Tom (LT): I'm a little confused, as I have used the links you commented out several times and they always worked for me. I'm pretty sure they have been used plenty of times. I'm guessing the Toolforge server was just down? Ovinus (talk) 17:51, 27 September 2020 (UTC)
Oh dear! I have tried the links on multiple articles over one to two weeks using multiple IP addresses, so that's really weird to hear. Can I confirm you have used these links in the last week or so? Perhaps a third editor can also try and see if those links are working. If they are, I am left feeling very confused.--Tom (LT) (talk) 06:09, 28 September 2020 (UTC)

Syncing WikiLove barnstars with local templates

mw:Extension:WikiLove is, apparently, commonly used for handing out barnstars. We also have our local barnstars, at Wikipedia:Barnstars#General_barnstars. Some of these are in sync with the extension, others are not (e.g. the extension uses label "The Technical Barnstar", whereas we locally use {{The da Vinci Barnstar}} (with alt option)). It's a very small issue, but just wondering if there's some neat way we can keep the two in sync? The extension unfortunately seems to use 'phrases' as pages, not templates (and MediaWiki:WikiLove.js as central config), so I don't think we can just create each MediaWiki page and point it to the template. It'd also be neat to give some control to community (or limit to TEs, if need be) to be able to add/remove/modify the barnstars shown. Also pinging @Kaldari and MSGJ: who might know. ProcrastinatingReader (talk) 20:04, 27 September 2020 (UTC)

@ProcrastinatingReader: It's easy to add new barnstars to WikiLove (regardless of whether they have a template or not). If you look at the MediaWiki:WikiLove.js page on MediaWiki.org you can see how they added a Translator Barnstar there. The MediaWiki:WikiLove.js page on Commons has more examples, adding the Tireless Contributor Barnstar and Copyright Watcher Barnstar. You can also delete barnstars if they aren't relevant here. (See the bottom of the Commons page for examples of deleting barnstars.) But yes, you do need Administrator or Interface Editor rights to modify the MediaWiki:WikiLove.js page. Kaldari (talk) 22:57, 27 September 2020 (UTC)
Kaldari, well, it can be edited by admins but I mean more in the sense that there's WikiLove extension barnstars that have the same purpose as our local ones, but with different names (eg "The Technical Barnstar" v "The da Vinci Barnstar"). I think it's a bit confusing (took me a while to figure out that the former was coming from an extension). And the limitation to sysops only, to edit, I think may make it used less (our WikiLove.js is far more empty than the Commons one). I think some way to have it in sync with our local barnstars, via a template, and also modifiable as such, would be an improvement. One thought I had was perhaps a template with a switch returning particular values to the MediaWiki pages, since the extension only exposes the phrases and not the full template for editing, but this wouldn't add the ability to add/modify barnstars. ProcrastinatingReader (talk) 14:11, 28 September 2020 (UTC)

Listing additions/removals of a category

Is there a tool that allows you to see recent additions and removals of a category? I recently realized you could make this if you used the Recentchanges API, but I wonder if there already is one. Nardog (talk) 14:22, 28 September 2020 (UTC)

@Nardog: add the category to your watchlist, then on your watchlist uncheck the "hide page categorization" box. — xaosflux Talk 14:26, 28 September 2020 (UTC)
@Xaosflux: I already do that, I'm asking if there's a way to list them per category without having to watch it because categories with frequent traffic can easily overflood the watchlist. Nardog (talk) 14:31, 28 September 2020 (UTC)
Go to Preferences → Recent changes, and disable "Hide categorization of pages"; and at Preferences → Watchlist, disable "Hide categorization of pages". --Redrose64 🌹 (talk) 14:29, 28 September 2020 (UTC)
@Redrose64: What I said to Xaosflux. Nardog (talk) 14:41, 28 September 2020 (UTC)
For a specific category, the categorymembers API can be used with sorting set by timestamp. For example, here's the 100 latest additions to Category:Living people. – SD0001 (talk) 16:06, 28 September 2020 (UTC)

Templates not rendering in ITN archives

Hello, is there a maximum number of templates per page? The ITN archives are large pages and recently we've noticed that the templates aren't expanded for all occurrences in the page. Consider the August 2020 archive. Someone said that there is an upper bound on the number of templates. Is this true? I didn't see anything obviously broken on that August 2020 page. --LaserLegs (talk) 20:05, 28 September 2020 (UTC)

@LaserLegs: Yes, WP:TLIMIT. --Redrose64 🌹 (talk) 20:11, 28 September 2020 (UTC)

Category:Pages with non-numeric formatnum arguments

Does anyone know how this category is being populated? In Marlborough, Calgary, it seems to come from the {{Infobox settlement}} parameter elevation_m = 1075, but I can't really figure out anything beyond that. kennethaw88 • talk 19:45, 24 September 2020 (UTC)

Well, the category page has been created now, but it was a redlink for a while, and it only had 250 members when I saw it first. It's still a mystery to me where it is being populated. kennethaw88 • talk 21:43, 24 September 2020 (UTC)
Category:Pages with non-numeric formatnum arguments is now hidden, which helps, but the issue seems to be in {{infobox_settlement/lengthdisp}}. I can't get it to behave at all — GhostInTheMachine talk to me 22:44, 24 September 2020 (UTC)
It's due to use of {{INR Convert|2.5|c}} but I can't investigate further at the moment. Johnuniq (talk) 23:56, 24 September 2020 (UTC)
The category is populated for example when formatnum is used with "1,234" instead of "1234" (I tried on User:Seudo/test). According to a user on the French Wikisource, this was not the case a few days ago. Besides, {{formatnum:1,234}} used to produce « 1 234 » and it now produces « 1,234 ». If I understand correctly, this category is useful because {{formatnum:1,234}} is incorrect anyway. But maybe someone could confirm that the behaviour of formatnum has changed recently... Seudo (talk) 21:57, 26 September 2020 (UTC)
This category was created by gerrit:626024, which was merged on the 15th, and deployed around the 24th. * Pppery * it has begun... 17:30, 27 September 2020 (UTC)
When formatnum is given a correctly formatted negative number, e.g. {{formatnum:−9000000}}, it assigns this error category. See T237467. This apparent false positive currently applies to {{US Census population}}, which shows population declines, so many articles transcluding that template appear in this category.
Does anyone know a different way to do what formatnum does, but that accepts negative numbers as input? It would be great to get articles transcluding {{US Census population}} out of this error category. – Jonesey95 (talk) 19:00, 28 September 2020 (UTC)
Of course the best thing to do is to fix formatnum, but a template can do this as a stop-gap:
{{#invoke:String|replace|{{formatnum:{{#invoke:String|replace|−9000000|^[—–−]|-||false}}}}|^-|−||false}} → −9,000,000
Yeah, crude and ugly ...
—Trappist the monk (talk) 19:13, 28 September 2020 (UTC)
That is just the sort of crude and ugly workaround I was hoping for. That way, I won't have to hold my breath for the WMF to fix this new formatnum issue. I have applied this workaround to {{US Census population}}, which should reduce the category's population by many thousands of articles. Thanks! – Jonesey95 (talk) 20:25, 28 September 2020 (UTC)
BTW, I have documented some fixes on the category page, and I think that I have fixed {{Infobox settlement}}, but there may be a lesser-used subtemplate that still needs a fix. I have not been able to puzzle out {{INRConvert}} yet; it has a lot of subtemplates. – Jonesey95 (talk) 20:27, 28 September 2020 (UTC)

Implement "smart" custom toggles which act based on their state

Do you know how long an enhancement request takes to be reviewed at Phabricator? I created a task, phab:T262622, more than two weeks ago and it still hasn't been commented. I don't know the workflow there, so if such change would likely take longer, would it be adequate to implement the request (given it's accepted) here (at the MediaWiki namespace)? The script linked in the task is my sandbox. I will change it sometime to just make it support a new behavior and not modify existing. Already posted something similar in WP:VPR but had no reply. Alexiscoutinho (talk) 16:13, 28 September 2020 (UTC)

@Alexiscoutinho: much like the encyclopedia content, there is no deadline for enhancement requests - they are done by volunteers. Asking for something is fine - but the only way to ensure that it actually gets done would be to do it yourself (write a patch, test it, upload it) - otherwise you have to wait to see if a volunteer developer is interested on your request. Sometimes these get worked on "quickly" (i.e. under a month), while there are some requests that have been open for over 10 years. — xaosflux Talk 16:26, 28 September 2020 (UTC)
@Xaosflux: Does this mean I can self-assign a task? Alexiscoutinho (talk) 16:29, 28 September 2020 (UTC)
Yes, you can self-assign a task, but you cannot self-merge (perhaps, obviously). Getting review time is a pain, AIUI, but if the case is good (see below comment), in the realm of possible. --Izno (talk) 17:00, 28 September 2020 (UTC)
@Alexiscoutinho: Phabricator tasks can sit for a long time, especially, as with your apparent problem article, when it's not obvious what the problem is or if the solution you have apparently devised is actually fixing the root of the problem. To pick on the specific, the COVID articles are in-general pathological in their assumptions about article size, rendering software, etc etc. Is adding to the already (for-Wikipedia) sizeable Javascript load these lines of software actually necessary for all articles? Especially when we have WP:MOSCOLLAPSE? --Izno (talk) 17:00, 28 September 2020 (UTC)
@Izno: I think my task description clearly explains the problem and even gives some examples to aid visualization. The solution/script I provided does fix the root of the problem although there is a less invasive way it can be implemented and which I will in the near future. I didn't want to say it was THE solution because I would expect more experienced coders to review it and maybe even find a better solution. If the change was implemented in MediaWiki core, it wouldn't increase too much the plugin's number of lines. If it was implemented temporarily in the MediaWiki namespace it would only boost the number of loaded JS lines in the pages that have {{Medical cases chart}}. I think WP:MOSCOLLAPSE doesn't apply here because the collapsing doesn't hide very important info, it just tidies the layout. Furthermore, no info would be omitted on mobile as collapsing is fully disabled in that case. Alexiscoutinho (talk) 17:33, 28 September 2020 (UTC)
I think my task description clearly explains As someone looking at it from the outside, no. --Izno (talk) 17:43, 28 September 2020 (UTC)
@Izno: Well then, when I update my script I'll try to rephrase the description. Alexiscoutinho (talk) 18:04, 28 September 2020 (UTC)
Yes, it should be recast. Perhaps in a user story template kind of fashion. --Izno (talk) 18:07, 28 September 2020 (UTC)
I think I have a basic idea what the Phab ticket is asking for, however following the scheme in mw:How to report a bug and structuring the task in separate sections is always welcome so others could also get a better grip. For general info on why tickets often remain open and uncommented, see mw:Bug management/Development prioritization. Thanks! --AKlapper (WMF) (talk) 18:40, 28 September 2020 (UTC)
@Alexiscoutinho: For better or worse, MediaWiki templates were never intended to handle such complex UI interactions. You're basically trying to build a suspension bridge out of matchsticks. Ideally, template output shouldn't be dependent on JavaScript at all, and should still provide all the relevant information even in a print version of the article. I would suggest that you consider simplifying things. Perhaps you could just have the article show the data for the past 15 days and forget about trying to maintain a complete archive of the historical data within the article. After all, there are plenty of websites out there for exploring COVID-19 data in depth. Wikipedia is an encyclopedia, not a medical data archive. Kaldari (talk) 21:02, 28 September 2020 (UTC)

Tech News: 2020-40

Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.

Recent changes

  • Admins can now see links to diffs of deleted revisions on Special:AbuseLog. This uses the interface of Special:Undelete. [26]
  • Editors are automatically added to some user groups. For example editors are added to autoconfirmed users when they have edited enough times and long enough. Abuse filters can hinder users from automatically getting user rights for a period of time. They can also remove rights user have. Wikis can now ask to change how long this period of time is for their wiki in Phabricator. It is currently five days. [27]

Problems

  • Last year some abuse filters stopped working because of a new change. If they tried to use variables that were unavailable for that action they would fail. This has now been fixed. [28]

Changes later this week

  •   The new version of MediaWiki will be on test wikis and MediaWiki.org from 29 September. It will be on non-Wikipedia wikis and some Wikipedias from 30 September. It will be on all wikis from 1 October (calendar).

Future changes

  • You can't see the language links to other language versions from the talk page or history page. They are also not shown when you edit an article. This could change. It is not decided if for example the history page should link to another history page or to the article. You can take part in the discussion in Phabricator.
  • The link colours could change. This is to make the difference between links and other text more clear. You can read more in Phabricator.
  • In your preferences you can choose to get different notifications on the web or by email. You will see Apps as one of the alternatives later this week. This is because the Android and iOS Wikipedia apps will use push notifications for those who want them. You can see the preferences on the test wiki. The goal is to have push notifications on Android in October and on iOS in early 2021. [29]
  • You can soon put pages on your watchlist for a limited time. This could be useful if you want to watch something for a shorter time but don't want it on your watchlist forever. It now works on mediawiki.org and will come to more wikis later. You can read more and see when it will come to other wikis.
  • You can see what Wikimedians think are the best new technical tools this year. You can also nominate them.

Tech news prepared by Tech News writers and posted by bot • Contribute • Translate • Get help • Give feedback • Subscribe or unsubscribe.

21:23, 28 September 2020 (UTC)

HasteurBot tasks carrying to other bots?

Sadly, the bot operator Hasteur passed away about 3 months ago. However, this meant that HasteurBot, which was therefore deactivated, could no longer help notify users of the pending G13 deletion, so that users would not have been able to have a 1-month period of working on the drafts in order to save it from G13. So is any bot operator willing to take over these tasks? Thank you. Eumat114 (Message) 01:59, 29 September 2020 (UTC)

User:MDanielsBot by User:Mdaniels5757 took over some of the tasks, including the one mentioned. But it doesn't seem to have run since 12 August. – SD0001 (talk) 10:58, 29 September 2020 (UTC) – SD0001 (talk) 10:59, 29 September 2020 (UTC)

Wikimail down?

This weekend I had two people use the "Email this user" function on my user page, only for me never to receive either email. (I knew they had emailed me because I received an alert in each case.) I could use their email functions to let them know I never received their email, & they could contact me by replying. I've checked my spam filter & they weren't there. (As far as I know the etherbunnies ate them.)

Anyone else encounter this problem? Are there steps to troubleshoot what the problem might be? (Maybe a new setting in the MediaWiki software was created & the result is to send any email I receive to the nearest burrow of hungry etherbunnies.) -- llywrch (talk) 06:10, 29 September 2020 (UTC)

@Llywrch: I've just sent you an email. Did you receive it? —[AlanM1 (talk)]— 09:03, 29 September 2020 (UTC)
A possibility is that your email entered into Wikipedia is not what you think it is. When was the last time you received a Wikipedia message? You could email someone you know (or me) with "email this user" and they could give hints about what your email address is. Johnuniq (talk) 10:51, 29 September 2020 (UTC)
Are you or the senders using Yahoo? --Izno (talk) 12:01, 29 September 2020 (UTC)
@AlanM1: no I did not. I did receive an alert you sent me a message.
@Johnuniq: the last time I received an email was earlier this month from someone at the Foundation asking me to participate in a survey. Before that, I received an email from another Wikipedian 4 September this year. And I am on the WikimediaUS-l mailing list, & last received an email from that list on the 20th. (I haven't looked to see if any messages have been sent to that list since then.)
@Izno: AFAIK, none of us use Yahoo. My ISP is a local independent business (I know, you're surprised any of those are still around), & I've been receiving email from other senders okay. (Although more recently I have to go thru the spam filter because some high-volume senders, like my daughters' school is being flagged as false positives.)
And although no one asked, yes I verified in my preferences that the email settings are correct. And no, I haven't changed any of the settings in the last month or so. -- llywrch (talk) 14:07, 29 September 2020 (UTC)
@Llywrch: Try emailing me. —[AlanM1 (talk)]— 14:22, 29 September 2020 (UTC)
I was able to email myself successfully with "Email this user". —[AlanM1 (talk)]— 14:31, 29 September 2020 (UTC)
Okay, what email provider do you use then? --Izno (talk) 14:44, 29 September 2020 (UTC)
You want me to publicly post here part of my email address? No offense, but that lowers my privacy a bit more than I'm comfortable with. (But if it helps you to troubleshoot, I can email you. The software allows me to email others, it just won't deliver to me. Not any more.) BTW, I tried the same feature on my account at commons earlier this morning, & that didn't work either. -- llywrch (talk) 17:06, 29 September 2020 (UTC)
@Llywrch: Hey, that's fine. I'm just trying to ascertain that your email provider is not one of the ones that we've had troubles with (hence why I asked about Yahoo). You should consider searching phab: for the name of the provider in question. If it's not there, either no-one else has had that problem with that provider or you're the first to complain about it.
The reason I think it is a provider issue is that I know that MediaWiki sends some headers in email that cause the email to get filtered out along the way without ever hitting your (spam) inbox (see for example previous issues with Yahoo), not because MW is doing anything bad but because the email provider is not handling the emails correctly. --Izno (talk) 17:36, 29 September 2020 (UTC)
If it is with my ISP, I'd honestly be surprised if anyone else encountered the same issue. I'm one of less than half-a-dozen Wikipedians/Wikimedians here in Portland, & I'd be surprised if one of them shared the same ISP as me. -- llywrch (talk) 21:43, 29 September 2020 (UTC)
I can confirm that at least one of the editors who tried to email Llywrch not only does not use Yahoo, but thought they'd gone bust years ago. I did get my copy of the email I sent, so it got that far in the system at least. When he realised he hadn't received the email, he emailed me via email this user, and I received that, and could reply to it, with no problem. DuncanHill (talk) 14:34, 29 September 2020 (UTC)
@DuncanHill: thought they'd gone bust years ago Surprise! :) --Izno (talk) 14:44, 29 September 2020 (UTC)
Note that "replying" to an email that you got from wikimail doesn't touch us at all - it is directly from you to the recipient (via your mail providers that is). — xaosflux Talk 14:56, 29 September 2020 (UTC)

Okay, now this is getting weirder. I went to AlanM1's page to email him, but that option no longer is visible to me. The only relevant options I see are "Block user", "Mute this user", & "Change user groups". I don't even understand how those would work, unless they are Admin options. (I can include a snippet of the screen shot if desired.) I'm using the legacy Vector skin, with refToolbarInstalled disabled -- but this shouldn't effect my receiving email; that's software unrelated to the Wikipedia web interface. -- llywrch (talk) 17:06, 29 September 2020 (UTC) P.S. I mean, the web interface shouldn't matter unless it is where my email address is stored... -- llywrch (talk) 17:09, 29 September 2020 (UTC)

@Llywrch: How about Special:EmailUser/AlanM1? —[AlanM1 (talk)]— 17:42, 29 September 2020 (UTC)
User:AlanM1 appears to be more e-mailable than User:Llywrch. When I click on Special:EmailUser/AlanM1 it *does* open a blank message form, suggesting that it ought to work, but the same thing does NOT function when I click on Special:EmailUser/Llywrch. When I do that for Llywrch it comes back with 'Enter username of recipient', and after the next step it says "This user has not specified a valid email address". Maybe there has recently been an update to the Wikimedia code that tries to check for valid email addresses. Llywrch could try troubleshooting by turning off the checkbox that says 'allow other users to email me' and then turning it on again. That ought to require a cycle of confirmation, and you can observe whether that confirmation works. EdJohnston (talk) 18:47, 29 September 2020 (UTC)
@EdJohnston: I'm guessing that Llywrch removed their email temporarily in the process of troubleshooting, as I was able to email them yesterday. —[AlanM1 (talk)]— 18:52, 29 September 2020 (UTC)
@AlanM1: Nope, I haven't changed that. (Although I am tempted to try that.) But I missed a system message I should have seen after I tried to email myself from commons, that due to multiple delivery failures, my email address has been unsubscribed." Just clicked the link, waiting on an answer. (And I tried the trick EdJohnston suggested, only to encounter an error message informing me that I needed an authenticated email address to use that feature. Am I in trouble because my account is so old -- I've had it for almost 18 years now -- that it dates before email addresses were required to be authenticated, mine never was, & now that someone has enforced that rule I am now locked out?) I'll contact my ISP & see if their mailserver reports anything. -- llywrch (talk) 21:43, 29 September 2020 (UTC)

@Llywrch: We know that something changed between yesterday 2020-09-29 02:02 UTC, when I was able to successfully (at least from appearances) send you an email, and today now, when it now says "This user has not specified a valid email address." If you go to Special:Preferences#mw-prefsection-personal, the last section looks like:

Email options
Email (optional):
foo@example.com
Change or remove email address
...

Does it show an email address for you? Have you tried clicking the change button? At this point, though, it might be best to get in touch with the login/authorization problem group, though (I can't immediately find a link; might be WP:ACC; don't sysops have some secret IRC channel they can use to ask the Borg?  ). I'm guessing either your email address got corrupted (somehow; unlikely), or maybe your ISP stopped accepting inbound mail from WP's SMTP server and it zapped or marked it invalid after a certain number of attempts. —[AlanM1 (talk)]— 22:18, 29 September 2020 (UTC) Updated —[AlanM1 (talk)]— 22:22, 29 September 2020 (UTC)

Part of the problem now is that my email address is now blocked because of too many unsuccessful deliveries. My ISP suggested I change my email to $USERNAME@mail.$ISP.org. Only problem is that I use my password so infrequently that I'm having problems remembering it to change my preferences. I'll update when I learn more. (PS I don't do IRC. Last time I tried to hang out in the regular IRC channel, everyone went silent while I was there. I don't do rejection well. :-) -- llywrch (talk) 23:48, 29 September 2020 (UTC)
OK, changing my email address to my ISP's mailserver fixed it. HIIK why all of a sudden the email address I've had for 10+ years suddenly was no good. (If anyone is curious, I did ask if the mailserver reported anything in their logs & can provide a bit more info as to what went wrong.) Thanks to everyone for their help! -- llywrch (talk) 00:03, 30 September 2020 (UTC)
Unless that's normally the way email addresses for that ISP are given, it seems less than ideal from a futureproofing standpoint, from what my gut remembers about configuring DNS and mailers. Anyway, I'm glad it's resolved. —[AlanM1 (talk)]— 00:40, 30 September 2020 (UTC)

Escaping a pipe character in system messages

Please can anyone advise on this issue. User:Awesome Aasim has requested that some system messages using {{fmbox}} are hardcoded as tables (like this) to avoid it breaking when a pipe character occurs. Hopefully he/she will be along to explain this better than I can. Is there any way that fmbox can be used successfully in these cases? — Martin (MSGJ · talk) 08:04, 22 September 2020 (UTC)

First, where is that breakage occurring, and how is it manifested? Second, this is sematically an improper use of a table element. --Redrose64 🌹 (talk) 08:32, 22 September 2020 (UTC)
The main issue is with the parameter $1 in system messages for the title blacklist. This is the parameter that is responsible for providing the regex that was tripped in the local or global blacklist. Because it can be frustrating for good-faith editors to know which filter tripped when they were creating a filter, this is more of a diagnostic that I suggested only be visible to extended-confirmed users. Unfortunately, because the regex sometimes includes a pipe character | it breaks many templates. I do not know if there is a way where this character can be accommodated without breaking any additional templates. If anyone more technical than me in MediaWiki knows, feel free to reply to this message.
(On a side note, this issue can easily be fixed if MediaWiki sent {{!}} instead to the parameter $1. This could maybe fixed in a bug request on Phabricator.) Aasim (talk) 08:39, 22 September 2020 (UTC)
$1 is not a regular template parameter, like {{{1}}}. Can this be solved by wrapping it in <nowiki>...</nowiki>? —⁠andrybak (talk) 10:11, 22 September 2020 (UTC)
@Awesome Aasim:, what do you think? — Martin (MSGJ · talk) 13:29, 25 September 2020 (UTC)
Can someone test the changes on test wiki first? If they work, then we can implement the nowiki change. Aasim (talk) 15:05, 25 September 2020 (UTC)
Awesome Aasim, judging by the page Wikipedia:What Test Wiki is not, it seems that https://testwiki.wiki/ should be used for such tests. —⁠andrybak (talk) 13:39, 30 September 2020 (UTC)

Gadget: "Require confirmation before performing rollback on mobile devices"

The 2nd gadget in browsing tab, titled "Require confirmation before performing rollback on mobile devices", doesn't seem to actually do anything for me. I think it is User:MusikAnimal/confirmationRollback-mobile by MusikAnimal. Is it just me or is it broken? If the latter, any other way to get rollback confirmations on mobile? The button is massive (many times the size of diff or something) and all over watchlist etc, I've hit it a few times myself accidentally. Thanks. ProcrastinatingReader (talk) 19:57, 27 September 2020 (UTC)

N.B. this gadget loads MediaWiki:Gadget-confirmationRollback-mobile.js — xaosflux Talk 13:38, 29 September 2020 (UTC)
Same here. It doesn’t work on minerva mobile. আফতাবুজ্জামান (talk) 19:00, 30 September 2020 (UTC)

AnomieBOT

AnomieBOT hasn't been running for a day or so, and a lot of important tasks are getting missed. The talk page says Anomie's on a break, so I don't know what the best course of action is here. Thanks, –Deacon Vorbis (carbon • videos) 15:29, 30 September 2020 (UTC)

And just like that, it's back 🤖. –Deacon Vorbis (carbon • videos) 16:44, 30 September 2020 (UTC)
For best results, it usually works to post on the bot's talk page, or the operator's talk page, and ping the operator. I posted at User talk:AnomieBOT, and the problem was fixed within 90 minutes. – Jonesey95 (talk) 19:01, 30 September 2020 (UTC)

reFill

Hello. Is reFill down? Not doing anything for me. Thank you very much. Caro7200 (talk) 19:25, 29 September 2020 (UTC)

Also not working for me - but was earlier in the day... GiantSnowman 21:36, 29 September 2020 (UTC)
Me, too. — Maile (talk) 21:40, 29 September 2020 (UTC)
And baby makes three...er, four. (Was fine earlier.) --Ser Amantio di NicolaoChe dicono a Signa?Lo dicono a Signa. 21:44, 29 September 2020 (UTC)
Yep it went down while I was in the middle of cleaning out the category. Citer is still working. It is good if the article has only a couple bare urls but if there is a bunch of them its worth waiting until refill gets going again. MarnetteD|Talk 22:04, 29 September 2020 (UTC)

Appears to be back, thank you. Caro7200 (talk) 21:03, 30 September 2020 (UTC)

Problem with Template:Self-reference tool

As can be seen at the bottom of Alien and Darien, {{self-reference tool}} does not always integrate itself into lists, instead creating its own <ul></ul> section. Per MOS:LISTGAP this is a bad thing. (It seems to work correctly at the top of lists, however: see Western for an example.) I suspect it's being caused by the comments in the template code, but I am not sure. Eman235/talk 22:55, 30 September 2020 (UTC)

@Eman235: Fixed in Special:Diff/981211097. Jackmcbarn (talk) 23:16, 30 September 2020 (UTC)
Thanks a lot! Eman235/talk 03:10, 1 October 2020 (UTC)

Maryland (not a bug, just curious)

Referring to WP:Teahouse#Maryland, I noticed that the third oldest edit to Maryland is showing a change of around 80 thousand bytes. I am curious to know whether this was because some edits were not accounted for or what kind of bug caused this, so could someone enlighten me on how this happens? 45.251.33.149 (talk) 15:59, 1 October 2020 (UTC)

See this recent thread. – Jonesey95 (talk) 16:09, 1 October 2020 (UTC)

Raleigh Memphis, Tennessee

I think the article name should be Raleigh, Memphis. An editor recently moved the page, and I'd fix it but there are a bunch of redirects. Thanks for your help! Magnolia677 (talk) 19:06, 1 October 2020 (UTC)

Category:Neighborhoods in Tennessee might be the norm to follow. This looks to have been originally named Raleigh, Memphis, Tennessee, as that is how it is in the Template: Memphis, Tennessee. It was redirected to the one above by Bneu2013 in April 2020. If there was anything on the discussion page before it was moved, I do not see it. — Maile (talk) 19:47, 1 October 2020 (UTC)

Get ready for horizontal scrollbars with Wikipedia's "New Look"

The Wikipedia "New Look" is coming: Wikimedia Blog: "New Look".

There are some nice features, like collapsible left sidebar. Then, there's the fixed-width content window, leading to horizontal scrollbars at browser widths less than the one they specify. "New Look" is currently active at French Wikipedia. Also at fr-wikt, eu-wiki, fa-wiki, he-wiki, pt-wiki. Mathglot (talk) 22:51, 23 September 2020 (UTC)

I hate to be grumpy and rude (no, really!), but it is clear to me that none of the people who worked on these "Desktop Improvements" has professional experience with UIs or usability. Looking at the Collapsible sidebar animation I see the sidebar disappearing and reappearing, with no reuse of the real estate it occupied. What good is that? According to that blog post, it's supposed to "improve usability and focus by allowing people to concentrate on the content itself"; I'd concentrate better without all that wasted space in my windows, which, in the collapsed case, is of absolutely no use to me at all.
The maximum line width "feature" appears to be similarly disrespectful of the user's needs, by forcing (what will probably appear to be arbitrary) wastages of space. I thought fixed-width design went out 10 years ago. Or was that only among clued professionals? Grumble, grumple, harrumpf, get off my lawn you consarned kids... — JohnFromPinckney (talk) 00:10, 24 September 2020 (UTC)
There's a professional UX designer and a UX engineer on the team working on this. I suspect they do, in fact, have professional experience with UIs/usability. Anyway, no projects except the volunteering early-adopter wikis are getting this before next year, per the FAQ. --Yair rand (talk) 00:57, 24 September 2020 (UTC)
Maximum line widths continue to be part of current web layout design (people still have trouble tracking text beyond a certain line length). Responsive web design will use techniques to rearrange page components appropriately as the screen size changes. (I haven't looked at the prototype changes, so have no comments on whether or not the design is responsive.) isaacl (talk) 02:27, 24 September 2020 (UTC)
Anyone with any experience trying to inspect now-you-see-me-comma-now-you-don't elements in a side-panel of their browser can attest that "responsive web design" is the antichrist. ―cobaltcigs 22:28, 27 September 2020 (UTC)
@JohnFromPinckney: I believe the justification for the fixed-width is that it's easier for people to continue reading from one line to the next when the text column isn't huge. This is why newspapers have all their text in columns since it would be difficult to go from one line to the next across a huge sheet of paper. The reason for the resurgence in fixed-width designs is due to the high resolution of monitors these days (creating the same problem as the newspapers' giant sheets of paper). I'm not a designer though, so please take all of my comments with a grain of salt. Kaldari (talk) 23:32, 27 September 2020 (UTC)
Apple's hi-res display on the iPhone also catalyzed responsive web design, as it was a very high resolution display but with a very small physical screen, breaking previous designer assumptions. As browsing increasing moved to phones, sidebars were replaced with bottom bars, and entries in menu bars became links to sections on one big scrolling page, as that design is readily adaptable to different screen sizes. isaacl (talk) 00:21, 28 September 2020 (UTC)
Newspapers have fixed-width columns because Linotype machines had a fixed slug width which could not be adjusted, although they were made in different widths so two different newspapers might have a different number of columns for the same paper size. The slugs were fairly narrow because the metal (lead) used to cast the slug would sag under its own weight if made too wide. --Redrose64 🌹 (talk) 12:08, 28 September 2020 (UTC)
Thanks, Kaldari. I have no doubt the reading of some material can be made easier with shorter lines of text. That's not my complaint (nor the complaint of the majority who've whined about it up at MediaWiki. Nevertheless, the standard response of the fine folks at MW seems to be, "but, no, the research shows..."). My complaint is, the implementation of fixed width prohibits me from using my computer the way I want to. If I want to widen my window (for example, to reduce the height of my window and see other info above/behind/below while still viewing the same content in WP), I should be able to productively do so. And if I want to more easily read some difficult text, I should be able to narrow my window, because I believe 100% that certain texts are more legible with shorter line-widths. The point is, I (and all other WP-users) should be able to use my/or devices as we see fit. We've got the power already, WM wants to take this flexibility away. — JohnFromPinckney (talk) 19:13, 28 September 2020 (UTC)
From what I understand, special pages, such as the watchlist and contribution pages do expand. That would certainly be useful for reading watchlists where someone has left a long edit summary. Dreamy Jazz talk to me | my contributions 19:28, 28 September 2020 (UTC)
I always put the web page inspector into its own window. isaacl (talk) 23:37, 27 September 2020 (UTC)
Kaldari, the change to a fixed width is great, in my view. It makes it much easier to place images and quote boxes. SarahSV (talk) 23:42, 27 September 2020 (UTC)
I'm not sure what the goal is here, but it seems like introducing a "known" screen width will encourage various "visual editor" fallacies. For example, it still would not be safe for editors to assume:
  • That sizing a floating image to exactly match the apparent "height" of the paragraph of text next to it (let's say 12 lines appears to be exactly 228px) makes it "safe" to float another image on the opposite side of the next paragraph (without having to worry about an ugly gap when the {{clear}} tag is used, or ugly "sandwiched" text when it isn't used),
  • That the 960px is the "correct" width for a panoramic photo of some city's skyline,
  • That a specific number of <gallery> images (perhaps six) will fit on one row (or worse, that knowing this magic number makes the relevance of each particular photo secondary to avoiding a remainder),
And we really don't want them to.
Consistent layout results would still be a lost cause in practice, because different clients will select different fonts to satisfy (vague) body { font-family: sans-serif; font-size: x-small; } CSS defaults, which may or may not be overridden on a per-user basis anyway. For consistent results, you'd have to hard-code it to an absolute font-size, provide the name of some font that everyone has, and probably a handful of other things like margins, padding, and line-height (someone please tell me what else I've forgotten) whose default values might differ from one client to the next. Obviously, the more specific you get, the more you alienate users who liked the way it looked before. Sure, you can always advise the hard of seeing to zoom in and out with ctrl+mouse-wheel in their browser, but that won't have consistent results either. Some programs will resize "everything" (and mess up gallery wrapping, plus attempts at full-width images) while others will resize "text only" (and mess up floating image layout). Others might do something in-between (test this, in fact).
I suppose the only foolproof solution would be to introduce a viewing mode which renders each article inside a scrolling PDF viewer box (similar to Google Books or Scribd, but with clickable links). This would introduce additional concerns about about vertical space and page-breaks, and generally suck for this type of content. But at least it would look the same for everyone. Achieving a deterministic "print-friendly" "perfect" "what-you-see-is-what-everybody-sees" layout may echo some aspirations of the "Book" namespace, for those interested in actually doing so (I'm not).
My strategy would be to avoid making any assumptions about the client side at all, except that every possible html output will absolutely look bad for somebody somewhere. This can be mitigated by using floating images more sparingly, and without trying to be so goddamn feng-shui clever about their arrangement. I figure if it matters that much, you probably have too many. ―cobaltcigs 07:22, 28 September 2020 (UTC)

This VPT section was intended more as just a notification of something going on at mediawiki. If you have comments that you would like people involved with the project to see or respond to, the project discussion page is here: mw:Talk:Reading/Web/Desktop Improvements. Mathglot (talk) 04:12, 24 September 2020 (UTC)

Thanks for sharing this, Mathglot. I did notice the team have said they plan to remove the minimum width soon, so the horizontal scrollbars on narrower windows should no longer be a problem. the wub "?!" 09:58, 25 September 2020 (UTC)
Browsing tests on the above-mentioned wikis suggest this only affects the vector skin (which, in my humble opinion, looks like donkey ass and was totally unusable even before this) and not monobook. So hopefully they'll leave it at that. ―cobaltcigs 22:28, 27 September 2020 (UTC)
From a technical perspective, we can add a gadget to remove what is the largest complaint (fixed width) - and if we really wanted to we could make it default... — xaosflux Talk 13:39, 28 September 2020 (UTC)
Some of it seems cooler than other stuff. The table of contents is quite interesting. Not sure the collapsible nav as implemented is an improvement, but as an idea it certainly is. Seems a bit weird that the width of the page does not resize as a result - leaves a bit of a void. Would make sense to have it collapsed by default though, probably. All of those links are useless to the average reader. ProcrastinatingReader (talk) 14:08, 28 September 2020 (UTC)
Other than the search thing I don't like any of these "improvements" but then again I've never liked change full stop. I've accidentally tried the new Vector and in the most politest way of saying this ... it's bloody awful!. As the saying goes: don't fix what isn't broken. –Davey2010Talk 20:36, 1 October 2020 (UTC)

Wikipedia. Has. Been. Very. Slow..... Lately

Is this just my crappy laptop or is there an issue with Wiki's servers? The pic upload on here and Wikicommons have also been particularly slow the past month or so. WisDom-UK (talk) 19:15, 1 October 2020 (UTC)

phab:T262239 - and the replag - Yes, it's been going on at least since early September. — Maile (talk) 20:39, 1 October 2020 (UTC)

Logged Out Forcefully

Tracked in Phabricator
Task T264370

I was logged out in the middle of browsing. Is there something amiss with the servers? -- Veggies (talk) 21:52, 1 October 2020 (UTC)

Just been logged out on desktop and phone. DuncanHill (talk) 21:53, 1 October 2020 (UTC)

This happened to me too, and my notifications were also briefly messed up when I logged back in. Jackmcbarn (talk) 21:57, 1 October 2020 (UTC)
Me too, just now - twice today, and once yesterday. — Maile (talk) 22:00, 1 October 2020 (UTC)
WP:ITSTHURSDAY, perhaps? Mz7 (talk) 22:04, 1 October 2020 (UTC)
Couldn't be bothered checking another device I was logged in and wasn't sure if I had somehow accidentally logged out or it was another case of the WMF killing sessions probably because of a bug so came to check. Nil Einne (talk) 22:13, 1 October 2020 (UTC)
  • phab:T264370. --Yair rand (talk) 22:14, 1 October 2020 (UTC)
    Also announced in Wikitech like last times [30] Nil Einne (talk) 22:18, 1 October 2020 (UTC)
    And just like the last times no announcement on en-wiki. DuncanHill (talk) 22:20, 1 October 2020 (UTC)
    It was clearly announced on an obscure message board's talk page, with no wikilinks to it, guarded by a tiger.. --Masem (t) 22:28, 1 October 2020 (UTC)
    Well the phab was opened 8 minutes after the start of this thread or 14 minutes before it was posted here or about 27 minutes ago. The Wikitech announcement was made about 24 minutes after the start of this thread or about 2 minutes before I posted it or about 12 minutes ago. Someone could come now to tell us what we already know, but I'm not sure I see the point. Frankly, I'm more concerned about the smaller wikis who may have fewer editors able to find out these things, and where many of the editors may not even be able to read either of those announcements. Nil Einne (talk) 22:29, 1 October 2020 (UTC)
    An IMO more interesting issue is that although "Incidents are documented at https://wikitech.wikimedia.org/wiki/Incident_documentation>", this seems to be the documentation for the June incident which I guess is maybe supposed to cover the July one too and now over 3 months later, all it does is link to the mailing list report and one Phabricator task. Nil Einne (talk) 22:40, 1 October 2020 (UTC)
    Another minor but IMO still more pertinent question is whether the logout really happened at 7:43 UTC and took ~14 hours to take effect, or it happened maybe at 21:43 UTC and someone put the wrong time (maybe from a calculation error) and/or time zone. Nil Einne (talk) 22:47, 1 October 2020 (UTC)
    I have asked that question on the phab thing. DuncanHill (talk) 23:14, 1 October 2020 (UTC)
    And in response the time has been changed from 7:43 UTC to 21:48 UTC. DuncanHill (talk) 23:47, 1 October 2020 (UTC)

Random article — exactly what does it do?

Hi, I'm interested in using Special:Random to collect statistics about articles. For example, what fraction are biographies, what fraction are US-centric, etc.. This is not for inclusion in articles (which would be OR) but to inform discussion and perhaps for external mention. I'm prepared to hit the button a couple of hundred times and examine what comes up, using standard statistical sampling analysis to assess the results. But this exercise would be futile if Special:Random provides a skewed selection of articles. Can it be assumed that each article has equal chance of being chosen? Answering this question probably requires precise technical knowledge of how it is implemented. Thanks. Zerotalk 14:20, 1 October 2020 (UTC)

@Zero0000: See GlobalFunctions::wfRandom. --Izno (talk) 14:34, 1 October 2020 (UTC)
From some very old notes: Each page is created with a random number in range [0,1) (from and including 0, up to but not including 1), called page_random and indexed. Special:Random chooses a random number and returns the next article with a higher page_random value. The result can be filtered ($wgExtraRandompageSQL) to try again if a specified condition is met. The default namespace is main (articles). You can append a namespace, for example: Special:Random/User. Johnuniq (talk) 00:54, 2 October 2020 (UTC)
Thanks to you both. The facility seems to satisfy my needs. Zerotalk 02:09, 2 October 2020 (UTC)
@Zero0000: Such statistics have been generated before, probably by some other method than a sample of random articles. See e.g. Wikipedia:Statistics, Systemic bias in Wikipedia Gender bias on Wikipedia#The gender gap in content for some. Personally, if I were to whip up some stats, I'd either use Template:Category count or make a Wikidata query. – Finnusertop (talk ⋅ contribs) 06:22, 2 October 2020 (UTC)

Strange User Page

How did this happen? NonsensicalSystem(err0r?)(.log) 13:01, 2 October 2020 (UTC)

It's a page with the title Editing User:Mungkorn789/sandbox. I see it's now been deleted. – SD0001 (talk) 13:23, 2 October 2020 (UTC)
It is in the Main namespace, so the title is rather evil but it is technically valid (see WP:NC-COLON and WP:NC-SLASH) and it seems that the "article" can just be created — see Editing User:GhostInTheMachine or Editing User:GhostInTheMachine/sandbox. Regardless, it has just been deleted. — GhostInTheMachine talk to me 13:26, 2 October 2020 (UTC)

Navbox funny display

I just noticed the navboxes at the bottom of Cijin District, Kaohsiung look rather strange on my browser. Can anyone else see what I mean? — Martin (MSGJ · talk) 21:09, 1 October 2020 (UTC)

The article was damaged by This edit about 9 hours ago. — GhostInTheMachine talk to me 21:36, 1 October 2020 (UTC)
Ganbaruby has just fixed it. — GhostInTheMachine talk to me 21:39, 1 October 2020 (UTC)
Woah, didn't even realize that I did. Nice. ◢  Ganbaruby!  (Say hi!) 21:48, 1 October 2020 (UTC)
I don't know what happened there either, but it's looking better now — Martin (MSGJ · talk) 07:15, 2 October 2020 (UTC)
@MSGJ: Most likely MediaWiki wasn't sure if the <li> from the note section was closed (there was a matching /li at the end) because it was mixed with wikitext list syntax, in which scenario it would 100% be due to the Remex processing changes from a few years ago. I expect that kind of parse failure would cause similar behavior to this case here. I suspect another fix would have been to be explicit about where the unordered lists started and stopped; hard to do with wikitext, but could have been done with the li and closing li. When he moved it to a refn template above, that made it clear to MediaWiki what was going on and corrected the navbox issue. --Izno (talk) 17:13, 2 October 2020 (UTC)
@SSastry (WMF): just as an FYI this happened; pretty sure it doesn't need explanation or fixing, but I'm sure you'd like to know about the anecdote ;). --Izno (talk) 17:16, 2 October 2020 (UTC)
@Izno: Thanks for the pointer! My brain is fried after a week of wikitext and html so I cannot immediately tell what happened there but glad it is all fixed. :-) SSastry (WMF) (talk) 21:30, 2 October 2020 (UTC)

Preferences and Watchlist not navigable after forced re-login

After Wikipedia logs me out (which it does periodically for some reason), when I re-log in several functionalities are...not functional for a shortish period. In Preferences, I can't switch to other tabs (e.g. Appearance, Gadgets): the buttons just don't work. On my Watchlist, the filter widget at the top just displays the three grey circles indicating it's loading, but it never does. The collapsible arrows for the nested changes lists do not expand and don't have the filled blue dot next to them. The last time this happened the issue mysteriously resolved after I inspected the page. I don't remember what I did previous times (other than switch from Chrome to FireFox). Anyone have suggestions for what the problem is? JoelleJay (talk) 23:59, 1 October 2020 (UTC)

My out-of-the-blue guess would be that your browser is caching some of the page components that mismatch with the primary page. Perhaps forcing your browser to reload all page components would help. isaacl (talk) 16:30, 2 October 2020 (UTC)
JoelleJay, Another thing to try in cases like this is to open an incognito window and log into that. That's largely the same as clearing your cache, but slightly less disruptive. -- RoySmith (talk) 20:29, 2 October 2020 (UTC)
Note that the instructions under the "Bypassing cache" section won't clear your cache, but just force your browser to reload the page components. isaacl (talk) 21:35, 2 October 2020 (UTC)

Double edit summary?

 

Lately I've noticed that on some - not all - edits, a double edit summary box appears. See example screenshot (from Japan). When editing a section or performing some other edit that would result in text appearing by default in the edit summary, both boxes are filled; however, only a summary entered in the second box is saved. Any idea why this is happening? I'm using Firefox 81 and Modern skin. Nikkimaria (talk) 15:57, 3 October 2020 (UTC)

@Nikkimaria: are you using WikiEd? Sounds like the issue under discussion at User talk:Cacycle/wikEd#wikEd bug report: two boxes for edit summary. Nthep (talk) 16:42, 3 October 2020 (UTC)
Yep, that'd be it, thanks. Nikkimaria (talk) 16:55, 3 October 2020 (UTC)

Wayback Machine no longer saves archives

  Resolved

On [31], it shows that there haven't been any archives since Sep 27, yet I have saved this page every day and when I go to the ".../web/*/..." page on those days that day's archive shows up. For example, see todays; it shows up now but won't tomorrow. What can I do to actually get these archives to work?  Nixinova   T   C   (ping to reply) 03:10, 2 October 2020 (UTC)

It's not just that page, every archive I've made over the past week or so doesn't work. I hope this is just a fetching issue and not that these archives no longer exist...  Nixinova   T   C   03:18, 2 October 2020 (UTC)
There is a new snapshot, taken within the last hour. A check of a few other sites implies that the WBM was not collecting (much? Anything?) in the last 4 days. — GhostInTheMachine talk to me 08:22, 2 October 2020 (UTC)
See, now that link does not show any saves on Oct 2. The links that worked yesterday don't today. This is very frustrating.  Nixinova   T   C   20:02, 2 October 2020 (UTC)
I've made a post on the Archive.org forums about this: [32]  Nixinova   T   C   06:53, 3 October 2020 (UTC)
And they're all back now, good.  Nixinova   T   C   01:46, 4 October 2020 (UTC)

Why no ping?

In this edit[33], I tried to ping to a list of editors.

However, subsequent comments at WP:Categories for discussion/Log/2020 October 3#Category:Arts show that most (maybe all) of the editors didn't get the ping.

I presume that I screwed up somehow, but I can't see an error.

Can anyone figure out what went wrong? I'd like to avoid repeating my error. --BrownHairedGirl (talk) • (contribs) 01:04, 4 October 2020 (UTC)

Because you also edited your previous comment, your change looked like an edit, not a new comment. The system tries to avoid pinging people twice and it uses a simple procedure of only pinging when a new comment is added. An alternative would have been to link to each user in the edit summary (cannot use a template for that). Johnuniq (talk) 01:49, 4 October 2020 (UTC)
Ah. Thanks, Johnuniq. I hadn't spotted that.
i will try in future to remember to do pings without also tweaking a previous comment. --BrownHairedGirl (talk) • (contribs) 04:34, 4 October 2020 (UTC)

Split links at cite template like harvnb and sfn into two different links

I propose that the sfn and harvnb template be divided into two links.

Currently, if we click a link in harvnb or sfn, say Foo (2016), p. 36, it will link into the book in the bibliography.

What if we split the links? Foo (2016) if clicked will link to the book in the bibliography, while p. 36 will link into the direct url of that page.

Google books supports these kind of directly linking into books. This link, books.google.com/books?id=I_IVAQAAMAAJ links into the book cover/the overview, while books.google.com/books?id=I_IVAQAAMAAJ&pg=PA20 links directly to the specific page. This is quite useful for old books which resides in google books. Regards, Jeromi Mikhael (marhata) 13:47, 3 October 2020 (UTC)

You can already do that with {{harvnb}}, {{sfn}}, and other related templates. Just wrap the page number in a link to the page at Google Books, archive.org, or another site, like this:[1][2] – Jonesey95 (talk) 14:19, 3 October 2020 (UTC)
Probably not fake? I requested for not fake. Regards, Jeromi Mikhael (marhata) 15:05, 3 October 2020 (UTC)
That doesn't look fake to me. What would you prefer that is flexible enough to handle a page link to any other website? --Izno (talk) 02:19, 4 October 2020 (UTC)
Sorry if the word "Fake" was misleading in the headers below. I have changed it to "Example". – Jonesey95 (talk) 05:48, 4 October 2020 (UTC)

Example bibliography section

  • Smith (2000). Title.

Example References section

References

  1. ^ Smith 2000, p. 20
  2. ^ Smith 2000, p. 21.

Copy-and-paste translation of source content of non-English page

Hello VP, there's problem of copy-and-paste wikitable content of Oteckovia in Slovak language to the List of Oteckovia episodes with full translation to English, that I failed to do action. Can you copy the Part section of the Slovak language of Oteckovia to the English Wikimedia list article by yourself? The Supermind (talk) 09:49, 4 October 2020 (UTC)

What should live preview do on user CSS/JS?

  Resolved

Even with the live preview on (Preferences → Editing →   Show previews without reloading the page), previewing a user CSS or JavaScript reloads the entire page just like when it's off. This is ostensibly to allow users to test the code as the "tip" above the edit box informs them, which is basically impossible without reloading the page. Its implementation, however, has a couple of problems:

  • Not just live preview but live diff is unavailable for user CSS/JS.
  • You can only test the code once. When you've already previewed the script once, the live preview takes over and simply shows you the new code without running it. You have to go back one page in order to test it again.

I thought I'd figured out how to fix these so I submitted a patch, only to be told by Ammarpad, the above patch does not actually fix the bug, it just worked it around by 'disabling the live preview feature'. In other words, it means ignoring users' preferences (for live preview) on user config pages at all. Which I found perplexing because the status quo already does go out of its way to "ignore users' preferences on user config pages" and the patch is all about doing it less (for live diff) and in a more consistent manner (reloading more than once). (I now know a better way to achieve it so you can abandon the patch now, which I don't seem authorized to.) Ammarpad also made me aware of an older task, T186390, that covered the same ground (which I'm grateful for) so I merged my task with it, but AFAICT that task also advocates for the same thing as the patch (it works the first time, but not afterwards ... live preview should just disable itself on JS pages). But since no one has further responded on Gerrit or Phab, I want to ask editors here before resubmitting the patch.

TL;DR: (1) Should live preview disable itself on user CSS/JS (as it currently does)? (2) Should it disable itself more than once (which it currently doesn't)? (3) Should live diff be available for user CSS/JS (which it currently isn't)? Nardog (talk) 09:32, 3 October 2020 (UTC)

Yeah, I think the option should be ignored on pages where it doesn't work. At least then people aren't surprised when it "stops working". We can leave the patch open to explain that there's probably more work that could be done to make the option actually work ina context I would not expect it to. --Izno (talk) 02:17, 4 October 2020 (UTC)
@Izno: Okay, I've updated the patch now. I hope it's clearer what I'm trying to do. Nardog (talk) 10:54, 4 October 2020 (UTC)

Parameter in ref tags

Hello everyone, I'm from ckbwiki. Today, we published three articles on ckbiwki by using a bot and ckb:حەلەب is one of them. Now, i want to know why the third parameter (it's article name on enwiki) doesn't show the article name on enwiki? It's printed in {{{3}}}. I think the ref tags are the problem. Can someone help? Thanks! ⇒ AramTalk 22:35, 29 September 2020 (UTC)

No one? Any idea? ⇒ AramTalk 13:49, 1 October 2020 (UTC)
Not the <ref>...</ref> tags but someone included {{{3}}} in the call to ckb:Template:بیرخستنەوەی ویکی. You can see that in the article source. I don't know what that template is but apparently it doesn't like it when |‏سەردێڕ‎={{{3}}} and |‏بەستەر‎=https://en.wikipedia.org/wiki/{{{3}}}. Give the template proper values and I suspect that you will get proper results.
—Trappist the monk (talk) 14:08, 1 October 2020 (UTC)
@Trappist the monk: Thank you for the response. See here. You can see the English article titles on the 3rd column from right and see the root page. Actually it works, but when I don't put the template between <ref>...</ref> tags, it's okay and {{{3}}} becomes the title of the English article, but when I put it between <ref>...</ref> tags, {{{3}}} does not become the title of the English article and it will be published as it is. The problem is not the template itself, but <ref>...</ref> tags don't let me get what I want. Is there any way to fix that? ⇒ AramTalk 20:16, 1 October 2020 (UTC)
I edited ckb:حەلەب, copied this from between the <ref>...</ref> tags:
{{بیرخستنەوەی ویکی |بەستەر=https://en.wikipedia.org/wiki/{{{3}}} |سەردێڕ={{{3}}} |زمان=ئینگلیزی |سەردان=٢٩ی ئەیلوولی ٢٠٢٠}}
Then I pasted it right at the top of the article, not inside <ref>...</ref> tags, clicked show preview and saw this (copied right off the rendering so there are no links):
ەشداربووانی ویکیپیدیا، «{{{3}}}»، ویکیپیدیای ئینگلیزی. سەردان لە ٢٩ی ئەیلوولی ٢٠٢٠.
So your claim that the 'problem' is due to <ref>...</ref> tags does not seem to be supported.
Pasting that same template call into ckb:Special:ExpandTemplates, I get this:
بەشداربووانی ویکیپیدیا، «<bdi><span class="plainlinks">[https://en.wikipedia.org/wiki/{{{3}}} {{{3}}}]</span></bdi>»، [[ویکیپیدیای ئینگلیزی]]. سەردان لە ٢٩ی ئەیلوولی ٢٠٢٠.
The problem, it seems to me, it that you are using positional parameters where you should be providing proper values for the named parameters. In article text, even in the call to a template, positional parameters like {{{3}}} have no meaning – they are just plain text which the template apparently uses to create link to en.wiki. But en.wiki does not have, nor does it allow, article titles that use the curly braces. Replace {{{3}}} with something meaningful (both instances) – I used 'Title' in the template call and the rendered template gave me a link to Title at en.wiki. Give ckb:Template:بیرخستنەوەی ویکی meaningful input and it should give you meaningful output.
—Trappist the monk (talk) 22:17, 1 October 2020 (UTC)
@Trappist the monk: Thank you very much for your responses! We fixed the problem by using this template. Thank you again! ⇒ AramTalk 15:15, 4 October 2020 (UTC)

404 errors in collapsible navigation menus gadget need fixing

MediaWiki:Gadget-CollapsibleNav.css (that's "Allow navigation menus to be collapsed" on Special:Preferences#mw-prefsection-gadgets) needs fixes, as the URLs to SVG files are 404 errors now. --Malyacko (talk) 15:56, 4 October 2020 (UTC)

@Malyacko: I will submit an edit request in the appropriate location. --Izno (talk) 16:16, 4 October 2020 (UTC)

Page Size tool has disappeared and pop-up referencing anchors aren't working either.

The page size tool has disappeared for me. It only seems to appear when editing a page. If I click it in that state it says "you need to preview the page" but when you do, the Page Size link disappears again. I'm on Chrome. Similarly the hover-over reference pop ups have stopped as well, and the search suggestion drop down is also inconsistently working. Darkwarriorblake / SEXY ACTION TALK PAGE! 16:56, 3 October 2020 (UTC)

Can you report what is in your browser console as described here? Ruslik_Zero 21:01, 3 October 2020 (UTC)

What is the behaviour you're seeing? - Page size link only appears when editing a page, not in normal viewing mode or preview mode, so you can't actually use it. What is the behaviour you're expecting to see? Click Page Size on left hand side under Tools and see the character count / page size of the current article How can the incorrect behaviour be reproduced? Describe exactly what you're doing and seeing. - Per above Confirm that you have tried bypassing your browser cache. - Yes Make note of the skin that you use, your browser, browser version, operating system, and operating system version. - Google Chrome is up to date Version 85.0.4183.121 (Official Build) (64-bit), Windows 10 19041.508 installed on September 20,2020

Append relevant JavaScript errors your browser logged. 

uncaught SyntaxError: Unexpected identifier
   at domEval (load.php?lang=en&modules=startup&only=scripts&raw=1&skin=vector:11)
   at load.php?lang=en&modules=startup&only=scripts&raw=1&skin=vector:17
load.php?lang=en&modules=ext.centralNotice.choiceData|jquery%2Coojs-ui-core|jquery.ui&skin=vector&version=1bbh1:149 JQMIGRATE: Migrate is installed with logging active, version 3.1.0
Back_to_the_Future:1 Unchecked runtime.lastError: The message port closed before a response was received.
DevTools failed to load SourceMap: Could not load content for chrome-extension://eoolfmmapnkhandljfaaofncecfakljd/dist/Versioning.js.map: HTTP error: status code 404, net::ERR_UNKNOWN_URL_SCHEME Darkwarriorblake / SEXY ACTION TALK PAGE! 22:44, 4 October 2020 (UTC)

Coolest Tool Award 2020: Call for nominations

The second edition of the m:Coolest Tool Award is looking for nominations (see announcement on wikimedia-l). Please submit your favorite tools by October 14, 2020. The awarded projects will be announced and showcased in a virtual ceremony in November. Thanks for your recommendations! -- for the 2020 Coolest Tool Academy: --JHernandez (WMF) (talk) 12:21, 5 October 2020 (UTC)

Tech News: 2020-41

Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.

Recent changes

  • There is a new tool where you can see which home wiki users have in discussions on Meta. This can help show which communities are not part of the discussion on wikis where we make decisions that affect many other wikis.
  • You can now thank users for file uploads or for changing the language of a page. [34]

Problems

  • There were many errors with the new MediaWiki version last week. The new version was rolled back. Updates that should have happened last week are late. [35]
  • Everyone was logged out. This was because a user reported being logged in to someone else's account. The problem should be fixed now. [36]
  •   Many pages have JavaScript errors. You can read more and now see a list of user scripts with errors.

Changes later this week

  •   The new version of MediaWiki will be on test wikis and MediaWiki.org from 6 October. It will be on non-Wikipedia wikis and some Wikipedias from 7 October. It will be on all wikis from 8 October (calendar).
  • Letters immediately after a link are shown as part of the link. For example the entire word in [[Child]]ren is linked. On Arabic wikis this works at both the start and end of a word. Previously on Arabic wikis numbers and other non-letter Unicode characters were shown as part of the link at the start of a word but not at the end. Now only Latin and Arabic letters will extend links on Arabic wikis. [37]

Future changes

Tech news prepared by Tech News writers and posted by bot • Contribute • Translate • Get help • Give feedback • Subscribe or unsubscribe.

16:24, 5 October 2020 (UTC)

@Jdlrobson: It might make more sense to put your list of errors on meta instead of here. (As it happens, none of the errors are from en for this goaround.) --Izno (talk) 17:19, 5 October 2020 (UTC)

url-access should come from Wikidata

Wikipedia has millions of refs, some without any url-access data and many with wrong information. Maybe this could be automated with Wikidata? --Palosirkka (talk) 07:38, 6 October 2020 (UTC)

How do I use parameter and link in a template?

Hello, I have created this template (which is a type of cite templates) on our project. Here, I have a problem. I want to give two values to the |فەرھەنگ= parameter, but as a result, one text (which is the English one) appears in the link. I used #switch and many magic words, but they didn't give me what I wanted. Can someone help? Thanks! ⇒ AramTalk 17:35, 6 October 2020 (UTC)

You can't pass multiple arguments to the same parameter. That's why many templates have e.g. |last1= (or simply |last=), |last2=, |last3=, and so on. Nardog (talk) 02:00, 7 October 2020 (UTC)

Adapting nested lists to mobile

Please take a look at a page with nested lists like this on your mobile phone.

Sub-bullet-points currently cause the text to move to the right so that it's barely readable on small screens, with only very little space for the text that is squished to the right.

Which solution do you suggest or prefer?

Nested lists probably have to adapt to the mobile display by getting displayed differently in the mobile view (en.m.wikipedia) at least if it detected a small screen or a mobile device in general. One possible solution would be simply displaying the nested header above the content (either view these sub-bullet-points or treestyle view that could look similar to or even use Template:Category tree). The mobile view is important and should not be neglected.

I created this entry as you may be interested in the main discussion here.

--Prototyperspective (talk) 10:12, 7 October 2020 (UTC)

edit requests at Talk:2020 Nagorno-Karabakh conflict

Could somebody take a look at this page? My browser outright refuses to load this talk page, I suspect due to either some form of code malformation or sheer length. Unsurprisingly due to the current armed conflict, there has been a flurry of edit requests and I can't try to answer any at this time. Thanks in advance to anyone that can help. Eggishorn (talk) (contrib) 18:30, 7 October 2020 (UTC)

@Eggishorn: the page is quite large, try loading it in mobile view. — xaosflux Talk 18:41, 7 October 2020 (UTC)
@Xaosflux:, thanks for the tip, I always forget that I use the mobile website address easily while on desktop. Unfortunately it means that I can't use tools like one-click archiver or edit request helper, though. I see from looking at the edit requests that long-time editors there seem to be adequately engaging with the requesters so I will just let them handle those. Thanks again. Eggishorn (talk) (contrib) 18:48, 7 October 2020 (UTC)
@Eggishorn: another option you could try is safemode - however this will also disable all gadgets and scripts, but it helps if one of those is the reason the page won't load. — xaosflux Talk 18:56, 7 October 2020 (UTC)

Accessing archived references

I know that this isn't, strictly speaking, a WP issue, but it arises for me here because this (WP) is the only place i try and access archive.org pages. If i click on this link [https://web.archive.org/web/20180717013125/https://www.indomiliter.com/panhard-vbl-penjaga-rumah-boediono/] in List of equipment of the Indonesian Army i get a Unable to connect Firefox can’t establish a connection to the server at web.archive.org message from my browser. This happens with every single archive link i try and verify (in every article i try, not just the one i mentioned), yet i know the links are active because i've used IABot to find them.. I will be ever so grateful if someone can tell me why this happens and what i need to do to stop it, because there are a lot of archived references i'd like to work with. I expect it's something very simple, silly even, so don't be worried about offending me by pointing out my complete lack of technical skill or understanding. Thanks in advance; happy days, LindsayHello 18:07, 1 October 2020 (UTC)

@LindsayH:: are you on a filtered internet connection? It's relatively common for strict DNS-based web filters to block web.archive.org, because unless your computer is configured to cooperate with the filter, it's impossible to tell which archived page you're looking at. Vahurzpu (talk) 18:17, 1 October 2020 (UTC)
Can you get to https://web.archive.org/ at all? If not, try it on a different device (like a phone or a computer), on a different internet connection (e.g. if you are having trouble on wireless, try a phone without wireless). Also try a different web browser (Chrome, Safari). – Jonesey95 (talk) 18:21, 1 October 2020 (UTC)
Anecdotally, I had that problem for a little while with one browser (though links worked in other browsers), then after awhile the problem went away. I surmised it was a browser problem that had gotten fixed on the back end. Schazjmd (talk) 18:24, 1 October 2020 (UTC)
Vahurzpu, i don't even know what a filtered internet connection is. My laptop is plugged in directly to my router which is plugged into my broadband and i don't think i've got any software limiting what i can do or where i can go.
Jonesey95, i cannot connect to https://web.archive.org/ at all, neither in Firefox nor in Chrome (Chrome says This site can’t be reached web.archive.org refused to connect, which i assume is essentially the same as the Firefox message). I can get to the site on my phone, which uses a wireless connexion to the same router; i guess i can use the phone to look up what i need and then type the entries into the browser on the laptop; less than ideal, but it should work. Thanks for the help and suggestions; happy days, LindsayHello 19:57, 1 October 2020 (UTC)
What operating system? --Izno (talk) 20:32, 1 October 2020 (UTC)
Windows 10, presumably the most recent version since MS does that cool "We're gonna update you whether you're ready or not" thing; Firefox 80.0.1. And the phone, which does work, is Android 9, which claims it is the current and up to date; happy days, LindsayHello 21:30, 1 October 2020 (UTC)
Lindsay, is it all of archive.org or just the web archives (I see you can't get to the subdomain above, but there are other parts of the site)? Do you live in a country with any marginally censorious regime? Have you tried uninstalling and reinstalling one of the browsers of interest (I'm wondering now if it is possibly one of your security certificates became corrupted; I don't know why that would impact just Archive.org). When your mobile can access it, is it connected wirelessly to your home router? I think the only other possibility pending/confirmed by those answers is that your ISP is blocking access. --Izno (talk) 02:47, 4 October 2020 (UTC)
Izno, it is not the whole of archive.org; i just went to https://archive.org/details/brooklynmuseum-o4368-the-great-red-dragon-and-the-woman for example, and it worked perfectly. I have uninstalled and then reinstalled Chrome, and still have exactly the same issue, the "refused to connect" message. Strangely, my phone won't now connect, so i don't know if it did previously or if i did something wrong ~ clicked an incorrect URL and thought it was on the archive site but not. I live in a country with no restrictions that i know of. My ISP is a very large, popular one ~ guaranteed i'm not the only person editing using it, probably not the only person on this page at the moment. You mention something about security certificates; is there something i can check or change there? I do appreciate the help; happy days, LindsayHello 06:58, 6 October 2020 (UTC)
We're past what I know on that point. One other thing I thought of was that you have some sort of ad or script blocking, but I'm pretty sure it's not that given you can access the other page. I am not sure who would be better to help with this. --Izno (talk) 16:23, 8 October 2020 (UTC)

How to do three brackets?

If there need to be single brackets and the first bracket is followed immediately by a Wikilink, is there a way to keep from having a space between them?— Vchimpanzee • talk • contributions • 18:12, 4 October 2020 (UTC)

[<nowiki/>[[foo]]] ([foo]) works. * Pppery * it has begun... 18:13, 4 October 2020 (UTC)
{{!(}} is documented to produce the open square bracket (just like {{Square bracket open}}) but it just uses the actual character so {{!(}}[[foo]] does not do what you want. That seems like a bug. Maybe it should use &#91; ?DMacks (talk) 16:13, 6 October 2020 (UTC)
Those templates are often used for creating links, tables, etc. where direct input doesn't allow it, so replacing them with numeric character references will probably break lots of templates. Nardog (talk) 08:31, 7 October 2020 (UTC)
A help desk question I just saw was about the same thing.— Vchimpanzee • talk • contributions • 17:58, 8 October 2020 (UTC)

How to show a live number of English Wikipedia articles on another Wikimedia project?

Hi all

I'm writing some instructions on Wikidata and would like to add a live number of articles on English Wikipedia to them. I know about {{NUMBEROFARTICLES}} which gives 6,817,806 but is it possible to get something similar to {{NUMBEROFARTICLES}} on another Wikimedia projects? Not sure if you can use templates across projects?

Thanks

John Cummings (talk) 14:32, 5 October 2020 (UTC)

No, you cannot access that information on a different wiki. --Izno (talk) 14:45, 5 October 2020 (UTC)
That's true, but {{NUMBEROF}} now shows data that is updated four times a day. The following shows values which I obtained just before posting this:
  • {{NUMBEROF|articles|fr|N}} → 2,254,469
  • {{NUMBEROFARTICLES}} → 2 254 493 (at frwiki)
GreenC has a bot which does the updating, and has updated many projects with the new NUMBEROF code. Johnuniq (talk) 23:05, 5 October 2020 (UTC)
Thanks very much @Johnuniq:, would it be possible for this to work on Wikidata to show the number of English Wikipedia articles? Currently the template doesn't exist. John Cummings (talk) 08:35, 8 October 2020 (UTC)
Yes, but I can't do it now. I'll ping GreenC again to see if they are available. Otherwise I'll look on the weekend or maybe sooner. Remind me if I forget. Johnuniq (talk) 09:08, 8 October 2020 (UTC)
Thanks you very much for all your help @Johnuniq:, really appreciate it. John Cummings (talk) 09:44, 8 October 2020 (UTC)
@John Cummings: now installed: https://www.wikidata.org/wiki/Template:NUMBEROF -- GreenC 18:16, 8 October 2020 (UTC)
@GreenC: thank you so much, really helpful, if there's anything non technical I could help with let me know. John Cummings (talk) 21:08, 8 October 2020 (UTC)

Cannot verify my email address

I have had my account for years. Several weeks ago, I received a Wikipedia alert which said, "Your registered email address, xxxx@xxxxx,com, has been unsubscribed due to multiple message delivery failures. You can verify your email address again." Two days later, I stopped receiving Wikipedia emails. I have attempted on several occasions to verify my email address, but I never receive any email back from Wikipedia. What can I do to remedy this?--Quisqualis (talk) 08:15, 8 October 2020 (UTC)

@Quisqualis: This may be related to this problem experienced by llywrch recently. —[AlanM1 (talk)]— 14:45, 8 October 2020 (UTC)
Thanks for the pointer, AlanM1. I found that by removing and then immediately replacing my email address, the problem of not being verified has been solved.--Quisqualis (talk) 22:44, 8 October 2020 (UTC)

Testing article issue detection AI

Help us test our problematic statement detection system. We hope to deploy this to help editors.

We are developing an AI to automatically detect issues in articles related to: NPOV, CLARIFY and CITE. We need help evaluating how well the model is working. We are asking for a group of volunteers to evaluate a set of sentences that are flagged by the AI. The landing page for evaluations can be found here. This page has a small set of examples for each issue. We have included sub-pages that include more examples (e.g. More POV examples). If you want to help, please assess as many example statements as you can. The more assessments we get, the better we can judge our model and make improvements. A description of our research project is on meta. Sumit (talk) 22:43, 8 October 2020 (UTC)

I'm working with Sumit on this project. You might remember some of the related work we did with mw:ORES vandalism detection and topic routing. We think this will also be a useful tool. Thanks in advance for your help! --EpochFail (talk • contribs) 22:48, 8 October 2020 (UTC)

Odd one: category exists or not?

Take a look at Category:Wikipedia sockpuppets of Viraj Alom. Each user's userpage is identical, unless I'm missing something, which I probably am, because User:Tamilreporter123 insists on telling me 

"Category:Wikipedia sockpuppets of Viraj Alom does not exist. Please Use this link to create the category page. (The page will be pre-loaded. All you need to do is save it)"

Puzzled. --jpgordon𝄢𝄆 𝄐𝄇 19:00, 8 October 2020 (UTC)

  Works for me @Jpgordon: are you still seeing ths? Could you reload the page / purge your cache? — xaosflux Talk 19:11, 8 October 2020 (UTC)
I'm not seeing any problem either. But, for future reference, there's also caching that's happening on the server side, and the symptom you describe is a classic example of incorrect server-side cache invalidation. You can fix that by going to Special:Purge. In this specific case, https://en.wikipedia.org/w/index.php?title=User:Tamilreporter123&action=purge would be what you want. -- RoySmith (talk) 19:18, 8 October 2020 (UTC)
Yeah, caching. I cleaned the cache on my side to no avail, but it's resolved itself since. --jpgordon𝄢𝄆 𝄐𝄇 22:55, 8 October 2020 (UTC)

Grant request for localising Twinkle to other wikis

I have requested a m:Rapid Grant to add internationalisation and localisation support to WP:Twinkle (which would make it easy to port the gadget to other wikis). Any comments and community endorsements would be welcome at the Grant request page. This task has been stalled for over 8 years. Now may be the time it actually happens! See GitHub issue for the detailed plans on the work to be done. Regards, – SD0001 (talk) 18:41, 9 October 2020 (UTC)

Searching for cases of multiple templates on one page

I there a tool, or some search wizardry, that I can use to find all articles that have both of two given templates on them?

For example, I know that Ba-ta-clan features both {{Offenbach operas}} and {{Jacques Offenbach}}. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:34, 9 October 2020 (UTC)

You can try with petscan. --Gonnym (talk) 14:43, 9 October 2020 (UTC)
@Pigsonthewing: this is the list of articles that transclude both of those templates:
List
Did that semi-manually by: Exporting the what-links-here transclusions, dumping them to a file, sorting, then using a regex to only keep duplicate lines: ^(.+)\r?\n(?!(.|\r?\n)*\1). — xaosflux Talk 14:45, 9 October 2020 (UTC)
Thank you, but that was a simple example; my use-case is for more highly-used templates. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:10, 9 October 2020 (UTC)
Special:Search. You may also use a minus symbol to identify pages without one or the other. --Izno (talk) 14:54, 9 October 2020 (UTC)
That's what I wanted. Thank you. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:10, 9 October 2020 (UTC)

I've hit a snag, in that I am getting results which include things like {{One source|section}} (see this example, which includes Pretoria#Jacaranda city). Is it possible to exclude such section-specific cases? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:39, 9 October 2020 (UTC)

@Pigsonthewing: using search, you can try a not-in-source argument as well, such as this. — xaosflux Talk 16:44, 9 October 2020 (UTC)
It should be hastemplate:"More citations needed" hastemplate:"One source" -insource:/\{\{One source\|section/ (the pipe in the regex has to be escaped). – SD0001 (talk) 18:38, 9 October 2020 (UTC)
You must consider redirects and whitespace also. --Izno (talk) 19:30, 9 October 2020 (UTC)

code error: out of place lines of text in the image of "Table of elements"

  Resolved

Description is there: Talk:Chemical_element#code_error:_out_of_place_lines_in_the_image_of_"Table_of_elements"
Steue (talk) 22:27, 9 October 2020 (UTC)

Technical volunteer needed for Arbitration Committee Elections

Due to an RfC update, the candidates page which typically randomizes the transclusion order of ArbCom candidates, should only randomize once for a given user and then stay fixed. The easiest way to implement this would be some form of hashing and then sorting by the hash, which will attain psuedo-randomness per user and then stay fixed for each user. This requires a Lua module, and I am not very experienced with Lua. If anyone is up to the task, please let me know, so I can work with you on getting this set up for this year's elections.—CYBERPOWER (Around) 14:08, 3 October 2020 (UTC)

I believe you are referring to the code at Wikipedia:Arbitration Committee Elections December 2019/Candidates#Standing candidates which appears as follows.
{{#invoke:random|list|separator=newline
|{{Wikipedia:Arbitration Committee Elections December 2019/Candidates/First candidate/Statement}}
|{{Wikipedia:Arbitration Committee Elections December 2019/Candidates/Second candidate/Statement}}
|{{Wikipedia:Arbitration Committee Elections December 2019/Candidates/Third candidate/Statement}}
|{{Wikipedia:Arbitration Committee Elections December 2019/Candidates/Fourth candidate/Statement}}
}}
That uses Module:Random which will render the list of statements in a different order each time the page is purged. There is no way Lua can make the list stay in a fixed order for each logged-in editor reading the page. I'm pretty sure Lua cannot discover the reader's user name, and even if it could, Lua has no mechanism to store per-user or any other kind of data. Some JavaScript might be able to handle that by putting it in the browser's local storage or a cookie, but I don't know if there is a precedent for a site-wide script to do something like that. The caching mechanism of MediaWiki is opposed to any procedure that attempts to show different text to different users. Johnuniq (talk) 23:21, 3 October 2020 (UTC)
@Johnuniq: Actually, there's one clever way it could be done. Much of the MediaWiki interface is re-rendered for each user, and within it, {{REVISIONUSER}} is the user who's looking at the page. We could seed the RNG with the username. The only tricky part is figuring out whether or not we have somewhere suitable in the interface where we can inject such code without affecting the rest of the site. Jackmcbarn (talk) 23:27, 3 October 2020 (UTC)
Thanks, I'd forgotten about the magic words. There is some confusion because Help:Magic words#Other variables by type has a note "This shows the last user to edit the page. There is no way to show the user viewing the page due to technical restrictions." Here is a test:
  • User = {{REVISIONUSER}} → User = Bruce1ee
Even if that worked, the next step would be tricky. Johnuniq (talk) 23:42, 3 October 2020 (UTC)
(edit conflict) Doesn't {{REVISIONUSER}} store the user who last edited the page, not the user who is currently viewing it? * Pppery * it has begun... 23:43, 3 October 2020 (UTC)
@Pppery: On regular pages, it's the user who last edited the page. But in parts of the interface, it is the user who's currently viewing it. Jackmcbarn (talk) 02:50, 4 October 2020 (UTC)
Correction: I'm almost positive this worked somewhere at some point, but now when I try it on testwiki, it just gives me the empty string. Back to the drawing board I guess. Jackmcbarn (talk) 02:56, 4 October 2020 (UTC)
For your first comment: I assume by "regular pages" you mean editable wiki pages, for that what you said is true, but the "interface" part is not correct. MediaWiki: namespace pages are considered part of the interface and they are just editable wiki pages. So the magic word will behave the same on MediaWiki:Example as it will do here. {{REVISIONUSER}} is based on revision. On virtual pages (such as SpecialPages where there are no revisions) it's null actually, and outputs empty string.

The only place where it's always the user doing the edit, is during 'edit previewing', that's before saving the actual edit, and that area is exclusive to the concerned user. Once it's saved, it's then either dynamic (shows the last revision user always), or it'll become hardcoded string username of the user who did the actual edit with it because of subst modifier. As far as I know, there's no magic word that refers to the current viewing user always. JS world does have some kind of that, not exactly though. It's called "relevant user".

For the second comment: You are probably thinking of {{subst:REVISIONUSER}}. – Ammarpad (talk) 16:52, 4 October 2020 (UTC)
As I commented on the RFC, this requires either a Javascript or a PHP implementation (Lua does not and will likely never know who is viewing a page), and I already discussed that probably neither option is a good other option. It is unfortunate my comment was not paid heed to. --Izno (talk) 02:12, 4 October 2020 (UTC)
Izno, I closed the RfC in favor of this implementation, because I believed the implementation of it was feasible. If it turns out it isn't, then I will have to amend the close. —CYBERPOWER (Chat) 22:38, 5 October 2020 (UTC)
I see the point of your opening comment about hashing now—using a hash of the current user's name concatenated with the candidate's user name as a sort key would suffice for a random order and would not require the storage of any per-user information. However, as Izno says, only PHP or Javascript can get the name of the user reading the page. PHP is not going to happen but Javascript is feasible. Except, I think the Javascript would have to be loaded for all users on all pages and that is also not going to happen. Accordingly, I think the current code (random order on each page purge) will have to do. Johnuniq (talk) 22:56, 5 October 2020 (UTC)
Cyberpower678 Hah, I was not chastising you. I know you did due diligence as closer. :) --Izno (talk) 23:00, 5 October 2020 (UTC)
Izno, :-) —CYBERPOWER (Around) 23:06, 5 October 2020 (UTC)
I thought everyone knew that this would require JS to implement. We could install it temporarily as a default-enabled gadget, couldn't we? Best, Kevin (alt of L235 · t · c) 00:55, 6 October 2020 (UTC)
If so, the gadget code should be wrapped in a loader that is only activated on the correct page(s), to avoid always loading the code on all page views. I can take a crack at writing the js this week(end) if no one beats me to it DannyS712 (talk) 05:29, 6 October 2020 (UTC)
DannyS712, the sooner, the better. We only have 2 weeks to have a working solution implemented. —CYBERPOWER (Chat) 15:22, 6 October 2020 (UTC)
...I have real life, sorry DannyS712 (talk) 15:49, 6 October 2020 (UTC)
@Cyberpower678: there are no deadlines -- just because the RfC said they want this, doesn't mean that it will be willed in to existence - is the fall back if this isn't created to go back to the prior year's random scheme? — xaosflux Talk 18:03, 8 October 2020 (UTC)
Xaosflux, sure it is, but I would ideally like to have the RFC implemented as requested if possible. —CYBERPOWER (Chat) 18:50, 8 October 2020 (UTC)
Wikipedia:Interface administrators' noticeboard/Archive 1 § Chess viewer has some discussion on how to implement a loader gadget that would load a gadget on demand for specific pages. User:קיפודנחש (aka kipod) may be interested in helping out, as this would be useful to implement the consensus reached at Wikipedia:Village pump (technical)/Archive 175 § Enable chess PGN viewer for chess articles. Also pinging User:Wugapodes. isaacl (talk) 16:31, 7 October 2020 (UTC)
will be delighted to assist. i do not have permissions, so currently can't help with actual implementation, but i will give any support i can, time permitting. the mechanism was originated at ruwiki (i believe). we adopted it in hewiki, and added support for gadgets. we wanted gadgets because RL is more efficient with them, e.g. by using minimized code. the mechanism is made of one template, say, {{Load script}}, and a dozen or so lines in common.js. for clarity, let me copy the js code here:
// On demand loading of scripts and gadgets, initial version from ruwiki.
// Detects uses of template "טען סקריפט"
// and loads specifically-named scripts or gadgets.
// for a script to be loadable this way, its name must begin with "Mediawiki:Scripts/"
// for a gadget, its name as defined in gadgets-definition must begin with "ondemand-"
if ( mw.config.get('wgCanonicalNamespace') !== 'Special' ) 
mw.hook( 'wikipage.content' ).add( function( content ) {
		var beenthere = {};
		$( '.executeJS', content ).each( function () {
			var script = $( this ).data( 'scriptname' );
			if ( script && $.trim( script ) ) {
				script = $.trim( script );
				if ( ! beenthere[script] )
					mw.loader.load( "/w/index.php?title=Mediawiki:Scripts/" + script + ".js&action=raw&ctype=text/javascript", "text/javascript" );
				beenthere[script] = true;
			}
			var gadget = $( this ).data( 'gadgetname' );
			if ( gadget && $.trim( gadget ) ) mw.loader.load( 'ext.gadget.ondemand-' + $.trim( gadget ) ); // np repetitions - resourceloader takes care
		} );
} )
so we scan the page for a special signature, ".executeJS", which is left by the template. the template "encodes" the name of the desired script, and the code above extracts it, and tries to load the script or gadget. we believe this is safe from exploitation - for instance, the can not be used to load just "any" gadget - only gadgets whose name begins with "ondemand-". those gadgets are hidden, and were deemed "safe" by the interface editors. similarly, only scripts whose name begins with "mediawiki:Scripts/" can be run this way, and again, this "namespace" is reserved for scripts approved by interface admins. let me know if there's anything i can help with. peace - קיפודנחש (aka kipod) (talk) 19:59, 7 October 2020 (UTC)
Thanks for the code and explanation. Dreamy Jazz talk to me | my contributions 20:28, 7 October 2020 (UTC)
@קיפודנחש: I found a way to exploit it: if the script name is &title=User:Mr._Stradivarius/evil.js&whatever=, then the URL passed to mw.loader.load will be /w/index.php?title=Mediawiki:Scripts/&title=User:Mr._Stradivarius/evil.js&whatever=.js&action=raw&ctype=text/javascript, meaning User:Mr. Stradivarius/evil.js will be loaded. (That page is a red link now, but can be created with a suitably evil script if desired. ;) — Mr. Stradivarius ♪ talk ♪ 12:19, 9 October 2020 (UTC)
@Mr. Stradivarius: - thanks. looking at it now. will ask our local expect to fortify it to prevent this specific explit. peace - קיפודנחש (aka kipod) (talk) 16:22, 9 October 2020 (UTC)
@קיפודנחש and ערן: It should be OK if you encode the script name using encodeURIComponent. I would encode the gadget name as well, just to be sure. That will help to protect against possible future changes to mw.loader.load itself. — Mr. Stradivarius ♪ talk ♪ 22:36, 9 October 2020 (UTC)
This use case is a far stretch from requiring adding javascript for every page view -- maybe a Gadget for rights=purge (to only target logged in users), but that is still a lot of overhead to force on every single page load, I brought up in the RfC that the desire to do something was being put in front of the work required to actually do it - so while the RfC can still say this is desirable, so are 10 years of backloged phab tickets - someone will want to really work on this. As far as page-specific scripts, yes I think that is a good idea, but shouldnt require running scripts on every page. Oh look, here is phab:T8883 from 2006 if someone really wants to work on per-page scripting! — xaosflux Talk 18:01, 8 October 2020 (UTC)
@Cyberpower678 do you have a mockup or example that I can use with to develop the script (i.e. a page transcluding multiple subpages that have different content, so I can test showing them in a random order)? DannyS712 (talk) 02:21, 9 October 2020 (UTC)
@DannyS712: I created a sandbox from last years: Wikipedia:Arbitration Committee Elections December 2019/Candidates/Sandbox, for the most part we'd expect the page to stay the same other than the {{#invoke:random|list|separator=newline section. — xaosflux Talk 12:58, 9 October 2020 (UTC)

Formatting ISO dates as per site/user preferences in JavaScript

The MediaWiki API returns timestamps in the ISO format (yyyy-mm-ddThh:mm:ssZ). Is there a module or something you can use in JavaScript that converts such a date according to whatever the site settings and user preferences on date format/timezone are, i.e. like the dates you see in page history, contributions, etc.? Nardog (talk) 09:08, 8 October 2020 (UTC)

Extended content
Try the #dateformat magic word — GhostInTheMachine talk to me 10:09, 8 October 2020 (UTC)
@GhostInTheMachine: Thanks, but that's far from what I'm looking for. It can't handle times (H:M:S), let alone timezones, does nothing if the user has no preference set (rather than use the site's default), has poor internationalization support ({{#dateformat:2020-10-08|ymd}} on zhwp returns "2020 10月 8", not "2020年10月8日 (四)", and {{#dateformat:2020-10-08|dmy}} on thwp "8 ตุลาคม 2020", not "8 ตุลาคม 2563"), and, above all, you'd have to make another API query to get the parsed result. Nardog (talk) 10:50, 8 October 2020 (UTC)
I generally use #time but that has a fixed format. The template {{time}} has more options — 02:27, 28 4月 2024 UTC / 02:27, 28 เมษายน 2024 UTC but it does not "officially" take a timestamp as input, although this sort-of works: 07:05, 2 11月 2018 UTC / 07:05, 2 พฤศจิกายน 2018 UTC. Sorry, not sure what else to suggest. Maybe post something at Template talk:Time? — GhostInTheMachine talk to me 11:32, 8 October 2020 (UTC)
Again, I'm looking for something ready-made I can use to convert ISO timestamps in accordance with the site locale/settings and user preferences in JavaScript (preferably without sending the timestamps to a server). I know I can probably use Moment or something and come up with an elaborate way to replicate what the MediaWiki interface does, I'm just asking if I'd be reinventing the wheel if I did so. Nardog (talk) 11:50, 8 October 2020 (UTC)
OK, for JS take a look at the CommentsInLocalTime gadget — it responds to local browser settings. You may be able to clone some of the code for your use. There is also a fork at User:Mxn/CommentsInLocalTime — GhostInTheMachine talk to me 12:30, 8 October 2020 (UTC)
For the nth time, I'm looking for a way to replicate the MW interface. That's not what that script does. Nardog (talk) 12:36, 8 October 2020 (UTC)
I don't know javascript, but this discussion seems relevant. More search results. – Jonesey95 (talk) 15:02, 8 October 2020 (UTC)
I think mw:ResourceLoader/Core modules#mw.user.options is the place, it just has basically no documentation as to the names of the options, much less which are available from Javascript. Krinkle Is that the place? --Izno (talk) 15:57, 8 October 2020 (UTC)
@Izno: mw.user.options.get('date') and mw.user.options.get('timecorrection') will give you the user preferences (like "dmy" or "ZoneInfo|-240|America/New_York"), but that's not enough information for a script to convert a date in the ISO format to "16:20, 8 October 2020" or what have you. Nardog (talk) 16:20, 8 October 2020 (UTC)
I'm not sure I understand what the issue is. Are you asserting that mw.user.options says "here are the options" and not "here is this user's selection"? I am otherwise fairly certain that you can implement what you're looking to since ISO format encodes the applicable time zone (which is Z, in yyyy-mm-ddThh:mm:ssZ, for Zulu time aka UTC). --Izno (talk) 19:32, 8 October 2020 (UTC)
Or is the part you're stuck on turning ISO into English? Date might help for that. Otherwise no, I do not think there is anything in the MediaWiki ecosystem that does that. --Izno (talk) 19:39, 8 October 2020 (UTC)
Again, what I'm looking for is something ready-made that automatically figures out the site settings and user preferences and turns an ISO timestamp (or Date object or Unix time) to the exact format you see in page history, contributions, etc., nothing else. Nardog (talk) 23:37, 8 October 2020 (UTC)
Otherwise no, I do not think there is anything in the MediaWiki ecosystem that does that. --Izno (talk) 00:29, 9 October 2020 (UTC)
I don't think the API even provides a way to identify the site-default date format used on the wiki. Also note that the timecorrection value you get from mw.user.options doesn't account for DST. In JavaScript gadgets, presenting such things like dates and numbers in an i18n/l10n-compatible way is just one big headache. – SD0001 (talk) 19:21, 9 October 2020 (UTC)
Thanks. It's a shame there's no access point to the date formatting for JS even though the formatted dates appear in so many parts of the interface. Nardog (talk) 12:20, 10 October 2020 (UTC)

Block log has talk page link when more than one IP is blocked

I was curious why someone with no contributions other than a Help Desk question would be blocked and I found this. There shouldn't be a talk page link for an IP that doesn't exist.— Vchimpanzee • talk • contributions • 22:01, 9 October 2020 (UTC)

Vchimpanzee, per the talk page history, it was created in error. I have tagged it for speedy deletion. Home Lander (talk) 22:15, 9 October 2020 (UTC)
Is there not some way to keep a link to the talk page from appearing?— Vchimpanzee • talk • contributions • 15:25, 10 October 2020 (UTC)

External links false positive

This edit produced the page that begins "Your edit includes new external links. These may be much welcomed links to references" and requires a CAPTCHA to be passed. That wasn't a problem for me, but I've seen it happen in the past, and I now note that that page does not link to any way to report false positives, comparable to WP:Edit_filter/False_positives/Reports. I suggest (1) that it should, and (2) and the error behind this false positive be corrected.

There is no reason to reply to me about this; please discuss it here if you want, or move this report to a better forum if you know of one, and either take the actions or not. Thanks for your attention. --174.89.48.182 (talk) 06:11, 10 October 2020 (UTC)

Possibly, there's a malformed link already on the page that the software thinks is "new" with every edit (as in phab:T223195). As no one seems interested in fixing that bug, probably best just to remove the link, though I don't want to hunt for it right now.
Perhaps the message should link to this page (WP:VPT)? Or would we end up with too much garbage from drive-by spammers? Suffusion of Yellow (talk) 06:23, 10 October 2020 (UTC)
The link to http://www.recordonline.com/apps/pbcs.dll/article is probably the problem -- it redirects about 7 times and then redirects in a circle — GhostInTheMachine talk to me 15:52, 10 October 2020 (UTC)
Actually, I couldn't reproduce this. So maybe a one-off glitch. That link shouldn't cause a problem; the servers don't visit any of the links before you save (though some bots might afterwards). Suffusion of Yellow (talk) 00:41, 11 October 2020 (UTC)

Extract a domain from a URL with Lua

I want to get the domain (like aberdeennews.com) from a URL (like http://articles.aberdeennews.com/2007-09-23/news/26395715_1_kosher-meat-aberdeen-congregation-member). mw:Extension:Scribunto/Lua reference manual#Patterns is nice enough to mention that character escaping is not supported (so I must make an ugly workaround for that), alternation is not supported (to make my life harder I presume), everything is greedy (to make my life a living hell I presume) and in the end, I just can't figure it out whatever I try.

The closest I got was {{#invoke:String|match|s={{#invoke:String|replace|source=http://articles.aberdeennews.com/2007-09-23/news/26395715_1_kosher-meat-aberdeen-congregation-member|pattern=.|replace=@|count=all|plain=true}}|pattern=([^/@]*)@([^@/]*)/|start=1|match=1|plain=false|nomatch=not_found}} (aberdeennews) and it doesn't work. (also it's very ugly) — Alexis Jazz (talk or ping me) 13:31, 10 October 2020 (UTC)

This can be done in Lua with mw.uri.new("http://articles.aberdeennews.com/2007-09-23/news/26395715_1_kosher-meat-aberdeen-congregation-member").host I can't find any Wikitext interface to that functionality, though. * Pppery * it has begun... 14:07, 10 October 2020 (UTC)
@Pppery: Thanks. I found a way now, but it's not very pretty. I'd prefer the function you provided, but I'm too close to snapping. I tried to create a small module with the code you gave me. I shouldn't have. I really shouldn't. I need to go do something else for the sake of my blood pressure. — Alexis Jazz (talk or ping me) 15:03, 10 October 2020 (UTC)
You can get the same result as mw.uri.new(url).host with string.match( url, "%/%/(.-)%/" ). In Lua % is the escape character and - is the only non-greedy quantifier.
So in a test function {{#invoke:Sandbox/Jts1882/Test|getdomain|http://articles.aberdeennews.com/2007-09-23/news/26395715_1_kosher-meat-aberdeen-congregation-member}} generates articles.aberdeennews.com|articles.aberdeennews.com (the result generated by the two methods separated by a pipe). —  Jts1882 | talk  16:25, 10 October 2020 (UTC)
As wikitext, {{#invoke:String|match|http://articles.aberdeennews.com/2007-09-23/news/26395715_1_kosher-meat-aberdeen-congregation-member|%/%/(.-)%/|plain=false}} -> articles.aberdeennews.com. * Pppery * it has begun... 20:30, 10 October 2020 (UTC)
@Jts1882 and Pppery: Thank you both! Good to learn about the %. However, I specifically need the domain (aberdeennews.com), not the subdomain. — Alexis Jazz (talk or ping me) 23:45, 10 October 2020 (UTC)
Does {{#invoke:String|match|articles.aberdeennews.com|[^.]-%.[^.]-$|plain=false}} -> aberdeennews.com work? * Pppery * it has begun... 00:15, 11 October 2020 (UTC)
You can try this to extract the last two components of the host name: {{#invoke:String|match|http://articles.aberdeennews.com/2007-09-23/news/26395715_1_kosher-meat-aberdeen-congregation-member|//.-([^./]+%.[^./]+)/|plain=false}} -> aberdeennews.com isaacl (talk) 00:25, 11 October 2020 (UTC)
@Pppery: That doesn't remove what comes after the TLD and doesn't work with http://aberdeennews.com/article.
@Isaacl: That appears to work, and looks much better! Thanks! — Alexis Jazz (talk or ping me) 00:38, 11 October 2020 (UTC)

If you are looking for .com domains, a string search will work. In general, however, it is not possible to extract the domain from a URL without a look-up table such as that described here. For example, the domain of aa.bb.com is bb.com but the domain of aa.co.uk is aa.co.uk. Consider the URL http://www.news.kku.ac.th/newspaper/content/view/545/86/ (from Christian Wolther). The host is www.news.kku.ac.th and the domain is kku.ac.th. Or, try http://www.portoacre.ac.gov.br/dados-municipio (from Porto Acre) which has domain portoacre.ac.gov.br. [I edited this comment to correct an error.] Johnuniq (talk) 03:58, 11 October 2020 (UTC)

Thanks for raising this aspect. I was going to comment but the page you linked to provides much more detail. isaacl (talk) 04:16, 11 October 2020 (UTC)

IP talk pages: how many?

How many user IP talk pages do we have? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:08, 10 October 2020 (UTC)

Could probably be answered by someone using Quarry, but I can't. — Alexis Jazz (talk or ping me) 00:28, 11 October 2020 (UTC)
@Pigsonthewing: 9,280,186 (quarry:query/48920). – SD0001 (talk) 13:00, 11 October 2020 (UTC)
Of course the count doesn't include the 500,000+ IP talk pages which MZMcBride deleted back in the day when such deletions were considered legit. – SD0001 (talk) 13:07, 11 October 2020 (UTC)

Very helpful, thank you. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:00, 11 October 2020 (UTC)

js edit needed to clear renamed category

  Resolved
 – The page was updated, just as with other protected pages an edit request would also suffice for such requests in the future. — xaosflux Talk 15:43, 11 October 2020 (UTC)

Category:Wikipedians who use IRC was renamed to Category:Wikipedians who use Internet Relay Chat; however the old name still has two entries due to code on User:Yamakiri/Userpage2.js. The user hasn't edited Wikipedia in six years so can someone with access sort out the changeover please? Thanks in advance. Timrollpickering (talk) 21:40, 10 October 2020 (UTC)

@Timrollpickering: I've added a note about this discussion at the Interface administrators' noticeboard. Graham87 05:50, 11 October 2020 (UTC)

That page contains zero valid javascript and should be renamed to something without the suffix .js. Clever trick to make his userpage content uneditable by others, but it probably shouldn't be allowed (for exactly this reason). ―cobaltcigs 08:58, 11 October 2020 (UTC)

Case-insensitive page names

I know that pages on Wikipedia are case sensitive, except for the first letter or the namespace prefix (which doesn't count as part of the page name.) For example, Wikipedia:Why create an account? and Wikipedia:WhY CreAte aN acCoUnT? are unequivalent. However, it seems that special pages break this rule. Ok, so "Special:" has a capital S and the first letter after the prefix needs to be capitalized. However, it seems that the entire page name of any special page is case-insensitive, for example: Special:CreateAccount and Special:CreAteacCoUNt are equivalent. Why is this? --Gioguch (talk) 22:41, 11 October 2020 (UTC)

Special pages are special :); moreover, they are not wikipages so the rules for wikipages do not apply. --Izno (talk) 23:07, 11 October 2020 (UTC)

Non-superscript refs for tables

Commonly, we often use tables with a single column set aside for references for the given row, but using the standard <ref> format, these refs come up as superscript and seem to waste space. Do we have a means to create a non-subscript version of a reference (otherwise conforming to all the same properties as a normal ref including elements like name and group)? --Masem (t) 15:50, 11 October 2020 (UTC)

Directly in the ref tag? No. Templatestyles can do it by removing the superscript, but I would advise against it generally, and this use case is not compelling to suggest otherwise. --Izno (talk) 23:04, 11 October 2020 (UTC)
You could do it like this:
Table showing direct reference links
Header1 References
Statement Smith 2006

Sources

  • Smith, Jane (2006). Title.
If that is compatible with the table you are using. – Jonesey95 (talk) 23:53, 11 October 2020 (UTC)

Signature not working

I recently changed my signature so that it has my name, and links to my talk page and contribs page. Here is my signature: ―[[User:Sportzpikachu|''Sportzpikachu'']] <sup>[[User_talk:Sportzpikachu|<span style="color:green">talk to me</span>]]</sup><sub style="position:relative;left:-4.4em">[[Special:Contributions/Sportzpikachu|<span style="color:green">contribs</span>]]</sub> It all works properly, but there is a gap between my signature and the time, as you can see: ―Sportzpikachu talk to mecontribs 07:12, 11 October 2020 (UTC) Any tips on how to improve/fix my signature?

Try this (based partly upon {{sfrac}} template):
―Sportzpikachu talk to mecontribs 08:43, 11 October 2020 (UTC)
―cobaltcigs 08:43, 11 October 2020 (UTC)
At 338 characters that would violate WP:SIGLEN. --Redrose64 🌹 (talk) 10:08, 11 October 2020 (UTC)
How about this? "―Sportzpikachu talk to mecontribs" ―[[User:Sportzpikachu|''Sportzpikachu'']] <sub style=display:inline-block>[[User_talk:Sportzpikachu|<span style=color:green;display:block>talk to me</span>]][[Special:Contribs/Sportzpikachu|<span style=color:green;display:block>contribs</span>]]</sub> It weighs in at 253 characters. Jackmcbarn (talk) 19:05, 11 October 2020 (UTC)
@Jackmcbarn: Seems to work, although I had to replace talk to me with my talk to make room for the quotes. ―Sportzpikachu my talkcontribs 00:14, 12 October 2020 (UTC)

Editing thumb style images via templatestyles

I'm trying to edit the basic display of thumb images using templatestyles. However I've not seemed to get it working. Am I using the wrong syntax on my styles.css page? See my test here:

Thanks in advance for any advice. T.Shafee(Evo&Evo)talk 01:39, 11 October 2020 (UTC)

@Evolution and evolvability: A few notes:
  1. It's the thumbinner class that has a border, not the thumb class, so if you want to recolor it, you have to use that one.
  2. To use .ns-0 or other classes on <body>, you need to use "body.ns-0" to keep MediaWiki from sticking ".mw-parser-output" on the wrong side of it and breaking it. (Ignore the warning that it's overqualified, see phab:T254671.)
  3. Using * and #content like you are is unnecessary.
  4. Since white and #FFFFFF are the same color, your specialization for ns-0 isn't accomplishing anything useful. To see an effect, change one of them to a different color.
I fixed all but the last point for you. Jackmcbarn (talk) 19:19, 11 October 2020 (UTC)
@Jackmcbarn: Aha, thank you. Exactly what I needed to get me on the right track. T.Shafee(Evo&Evo)talk 11:14, 12 October 2020 (UTC)

Browser jumps to edit box when editing

I have noticed this happening since probably late July. Whenever I try to preview a page that I am editing, my browser always jumps to the edit box. Since the default preferences for previews is above the edit box, my browser will jump to near the bottom of the page which will vary depending on how long the article you are previewing is. I noticed this also happens in other languages Wikipedia and on Commons, Meta but curiously, not on MW:MediaWiki. Does anyone know how to disable this? I want to start looking at the preview without concerning whether the page has done loading or not. It's also getting annoying when my Internet is slow so it becomes a surprise when I am reading while waiting for the page to complete but suddenly interrupted and get teleported to the bottom of the page. I have looked in preferences but couldn't find anything relevant other than setting so the preview appears below the edit box. But it will only make me jump to the top instead. RXerself (talk) 23:36, 11 October 2020 (UTC)

I have noticed that this only happens if I use the 2010 wikitext editor toolbar. It doesn't happen if I disabled it. It also doesn't happen when I'm previewing my common.js page. I have used the toolbar for years I don't remember it behaving this way. RXerself (talk) 07:30, 12 October 2020 (UTC)
Update: it looks like it only happen when I turn on the "Syntax highlighting" tool on the main namespace. If I turn it off it doesn't happen. If I am editing a subpage or a template it also doesn't happen even if the tool is active. RXerself (talk) 12:50, 12 October 2020 (UTC)

WikiProject on open proxies discussion

I have started a discussion at Wikipedia talk:WikiProject on open proxies which may be of interest to some of the technical folks here. The thread is Wikipedia talk:WikiProject on open proxies#Reboot. GeneralNotability (talk) 14:05, 12 October 2020 (UTC)

Linktrailing rules

Tracked in Phabricator
Task T47126

It is well known that the wikitext [[Foo]]bar produces a link that is six letters long: Foobar. But if the first letter after the closing braces is capitalised [[Foo]]Bar, the link is only three letters: FooBar. How long has this difference in behaviour existed? I first noticed it at Wikipedia talk:Criteria for speedy deletion#File PROD vs dfu (WP:F7d) in the post made at 13:26, 11 October 2020 (UTC) by Davidwr (talk · contribs) which contains the link WP:REFUNDED. --Redrose64 🌹 (talk) 12:41, 12 October 2020 (UTC)

@Redrose64: looks like it has been for at least 7 years according to phab:T47126 - notably it only matches [a-z] and fails on other things too like Foo2 , Fooä , etc. We wouldn't want it to trail on every character (especially not punctuation) - much more reading on it in the tracked phab ticket. — xaosflux Talk 14:24, 12 October 2020 (UTC)

Marking default gadgets

Hi all, I can't recall the old discussions on this - but wanted to get any input. This tech proposal is to include an indicator on gadgets that are enabled by default - so that if someone starts making preference changes they would more easily be able to revert to the defaults without having to erase all of their preferences. We've done this for a while on meta-wiki and you can see the inline indicators on your pref's page there. I've added a mock up to the default gadget for adding confirmation to mobile rollback, see in the second line of Special:Preferences#mw-prefsection-gadgets. Any feedback before updating other descriptions? Thanks! — xaosflux Talk 14:18, 12 October 2020 (UTC)

Do you mean this discussion back in July that resulted in Phab:T70689? — GhostInTheMachine talk to me 14:30, 12 October 2020 (UTC)
Great idea, but I think small text (with a trailing space) instead of superscript would be less ugly/confusing. Nardog (talk) 14:45, 12 October 2020 (UTC)
Meta shows a superscript "d" within the gadget description with a tooltip reading "Enabled for everyone by default". This does work and will help, but it is not very visible. Perhaps a larger label such as Default or Default or a big, bold blob  • of some sort? Just adding a plain text (Default) at the end of the description would also be good. — GhostInTheMachine talk to me 14:56, 12 October 2020 (UTC)
Or what about just an asterisk or a dagger or something at the end of each line, with a note at the top saying "* denotes gadgets enabled by default"? Nardog (talk) 15:02, 12 October 2020 (UTC)
Or even better, put these at the start of the line (with non-default lines padded) so that they display in a tidy column. That way, even a small mark will stand out. — GhostInTheMachine talk to me 16:31, 12 October 2020 (UTC)

Error message should not appear?

Not really a question about Wikipedia, but when you link to an invalid project (e.g. n:simple: for Simple English Wikinews), it will take you to Incubator. Using an unapproved language code (e.g. zxx and going to https://zxx.wikipedia.org will show an error message not produced by the WMF.) This should not be happening with existing codes, but testwiki:n: (Test Wikinews, not an existing project, takes you to a "This site can't be reached" error message not by WMF. IT should be taking you to incubator:Wn/testwiki or https://en.wikinews.org/wiki/Main_Page, but it's not and I believe this shouldn't be happening in local projects. Can a phab ticket help? --Gioguch (talk) 16:44, 12 October 2020 (UTC)

@Gioguch: This might be better suited for meta:Tech if it is unrelated to Wikipedia. (When would you write testwiki:n: instead of for example n:testwiki, and which resulting URL would you expect for each of them? I'm asking as your previous example n:simple: followed the "first language code, then site family" scheme. If I understand correctly... a clearer list of what you enter and which full URL you'd expect might help, to avoid misunderstandings.) --AKlapper (WMF) (talk) 16:59, 12 October 2020 (UTC)
@AKlapper (WMF): So I expected that writing testwiki:n: should give you the URL https://en.wikinews.org/wiki/Main_Page or https://incubator.wikimedia.org/wiki/Incubator:Main_Page. Instead, it gives https://test.wikinews.org/wiki/ which is an invalid URL. Invalid URLs shouldn't occur in Any WMF projects. Why is this happening? --Gioguch (talk) 17:16, 12 October 2020 (UTC)
@Gioguch: Please explain what your expectations are based on. I don't see why writing testwiki:n: should ever link to an English (!) Wikinews website. Maybe you meant n:testwiki: instead? Anyway, for further details, see meta:Help:Interwiki linking. --AKlapper (WMF) (talk) 17:41, 12 October 2020 (UTC)
So testwiki:n: does link to exactly where I expect it to (https://test.wikipedia.org/wiki/n:) - however as soon as you land there you get the n: redirect -- fixing that is beyond the scope of English Wikipedia settings though - suggest follow up on metawiki or phab. — xaosflux Talk 17:46, 12 October 2020 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Village_pump_(technical)/Archive_184&oldid=1090173209"