Wikipedia talk:WikiProject Check Wikipedia/Archive 6

Archive 1Archive 4Archive 5Archive 6Archive 7Archive 8Archive 9


WikiProject X is live!

 

Hello everyone!

You may have received a message from me earlier asking you to comment on my WikiProject X proposal. The good news is that WikiProject X is now live! In our first phase, we are focusing on research. At this time, we are looking for people to share their experiences with WikiProjects: good, bad, or neutral. We are also looking for WikiProjects that may be interested in trying out new tools and layouts that will make participating easier and projects easier to maintain. If you or your WikiProject are interested, check us out! Note that this is an opt-in program; no WikiProject will be required to change anything against its wishes. Please let me know if you have any questions. Thank you!

Note: To receive additional notifications about WikiProject X on this talk page, please add this page to Wikipedia:WikiProject X/Newsletter. Otherwise, this will be the last notification sent about WikiProject X.

Harej (talk) 16:57, 14 January 2015 (UTC)

Stripping long article names

  Resolved

Hello,

some long article names are stripped on checkwiki pages which makes fixing them much harder. Example: http://tools.wmflabs.org/checkwiki/cgi-bin/checkwiki.cgi?project=plwiki&view=only&id=83 - please have a look. ToSter (talk) 10:52, 18 January 2015 (UTC)

ToSter In theory, I've fixed the problem. Won't know for sure until next dump. Bgwhite (talk) 21:03, 22 January 2015 (UTC)
Bgwhite Thanks, we'll see. ToSter (talk) 19:26, 24 January 2015 (UTC)

Extend the scope of error #2?

  Resolved

Maybe we should extend the scope of error #2 from "break tag with incorrect syntax" to "tag with incorrect syntax"? -- Magioladitis (talk) 08:26, 24 January 2015 (UTC)

Error #16

  Resolved

@Bgwhite and NicoV: I think we can add soft hyphen (00AD) in our list. -- Magioladitis (talk) 09:33, 6 January 2015 (UTC)

Part 1 done. Bgwhite added this in CHECKWIKI's code. -- Magioladitis (talk) 08:53, 17 January 2015 (UTC)

Should be working with WPC also. --NicoV (Talk on frwiki) 08:16, 29 January 2015 (UTC)

Updating cawiki (and others)

  Resolved

Since the move to wmflabs, cawiki got updated only once a month. During the December update, something went wrong, and the whole list was screwed up, so no cleanup has been done for more than a month now. I figured we should wait for this month's update, but it's not coming, and I have noticed that several other wikis (whose list wasn't screwed up to begin with) are getting updated. Can you please update cawiki more often? --Joutbis (talk) 17:48, 28 January 2015 (UTC)

Joutbis cawiki was never updated before the move to wmflabs. cawiki is updated when a dump is available, which is once a month. Here is a listing of the most recent dumps. Next dump should be available in the next couple of days. Bgwhite (talk) 18:47, 28 January 2015 (UTC)
What do you mean by never updated? It has always been in CHVP like the other wikis; anyway, I'll wait. Thanks.--Joutbis (talk) 21:31, 28 January 2015 (UTC)
Joutbis Toolserver had stopped updating from dumps. It was scanning a very small subset of updated articles. The number of articles with errors dramatically increased when it moved to wmflabs. They (whoever they are) are supposed to do bimonthly dumps for all of languages except for enwiki. It was, more or less, doing that until they updated their equipment. See meta:Data dumps#Summary. Bgwhite (talk) 22:16, 28 January 2015 (UTC)

#85 additions

  Resolved

@Bgwhite and NicoV: Checkwiki now finds cases of empty gallery and center tags. -- Magioladitis (talk) 08:19, 1 February 2015 (UTC)

AWB removes empty gallery and center tags. -- Magioladitis (talk) 10:32, 1 February 2015 (UTC)
WPC detects them also. --NicoV (Talk on frwiki) 11:52, 1 February 2015 (UTC)

Extending #85 for empty unnamed ref tags ?

  Done

Hi, I wonder if we should extend #85 to look also for empty unnamed ref tags (like in this buggy VE edit). --NicoV (Talk on frwiki) 10:13, 19 March 2015 (UTC)

NicoV, a listing for frwiki can be found at User:Bgwhite/Sandbox2. It was generated with an AWB database scanner search, so these contain cases where the empty ref tags could be in nowiki tags. I'm currently doing a checkwiki scan for enwiki. If everything looks ok, will add it. Bgwhite (talk) 20:22, 20 March 2015 (UTC)
For some reason there are always several examples (without nowiki) on the Category:Pages with incorrect ref formatting. I guess there is a problem with the placement of the option on the wiki-insert panel. Doesn't really explain why people commit a change like it. Unbuttered parsnip (talk) mytime= Sat 05:35, wikitime= 21:35, 20 March 2015 (UTC)
Thank you Unbuttered Parsnip for telling me about the category. I got some test cases from the category. I was wondering why CheckWiki couldn't find any examples because people are fixing them in that category.
NicoV It has been added. Bgwhite (talk) 18:11, 21 March 2015 (UTC)
"people fixing them" = me mostly! -- Unbuttered parsnip (talk) mytime= Sun 05:22, wikitime= 21:22, 21 March 2015 (UTC)

Template calls with duplicate arguments

  Resolved

Hi, I've added #524 to WPCleaner to detect template calls with duplicate arguments (several arguments with the same name). On frwiki, this allows to work on the pages in fr:Catégorie:Page utilisant des arguments dupliqués dans les appels de modèle. --NicoV (Talk on frwiki) 23:07, 10 March 2015 (UTC)

Equivalent category on enwiki is Category:Pages using duplicate arguments in template calls, currently 44k pages in it if someone is looking for work ;-) --NicoV (Talk on frwiki) 10:23, 11 March 2015 (UTC)
NicoV, the person who would enjoy this is Frietjes. The cat is already on one that she watches. I wonder if there are many cases where the values of the duplicate arguments are identical? If so, have WPCleaner fix this automatically in bot mode. Bgwhite (talk) 05:14, 16 March 2015 (UTC)
Done, when values are identical WPCleaner will remove the first argument automatically. --NicoV (Talk on frwiki) 17:31, 16 March 2015 (UTC)
I wonder also if I should remove the first argument automatically when it's empty... --NicoV (Talk on frwiki) 18:04, 17 March 2015 (UTC)
NicoV, on this wiki, we have Wikipedia:Bots/Requests for approval/SporkBot 5 which does both automatically, with some exceptions (see discussion concerning scores and other parameters with numeric suffixes). Frietjes (talk) 18:10, 17 March 2015 (UTC)
Frietjes, thanks for the link, very interesting :-) I modified WPCleaner to do automatic replacements only if the argument name doesn't end with a digit (should cover the exceptions), both for equal arguments or first argument empty. --NicoV (Talk on frwiki) 21:10, 17 March 2015 (UTC)

Update:

  • Frietjes Automatic replacement is also prevented if the argument name contains a digit (even if it's in the middle of the argument name) to avoid this (correct modification)
  • I'm starting to make errors > 500 behave more like regular CW errors. For example, defining the parameter category for #524 allows the bot tools to work on the list of pages for #524 as if it was available on labs like regular errors. I'm going to extend this behavior to other errors and also outside of the bot tools.

--NicoV (Talk on frwiki) 08:45, 26 March 2015 (UTC)

That's great! Would it be problem to make it easier by checking Special:TrackingCategories (in this case MediaWiki:Duplicate-args-category)? Matěj Suchánek (talk) 17:30, 26 March 2015 (UTC)
Thanks Matěj Suchánek, I will try to use it. --NicoV (Talk on frwiki) 06:35, 27 March 2015 (UTC)
Matěj Suchánek, WPCleaner is now checking MediaWiki:Duplicate-args-category if no configuration is provided. --NicoV (Talk on frwiki) 00:37, 28 March 2015 (UTC)

Typo

  Resolved

The main page currently reads " ... facilitates the correction of detected by the program ... ".

This seems to be a typo? Trafford09 (talk) 00:37, 23 February 2015 (UTC)

Thanks Trafford09, I've added "problems". --NicoV (Talk on frwiki) 10:37, 26 February 2015 (UTC)

When to start the check process

  Won't fix

I have noticed that the Check Wikipedia process runs the day after the monthly dump is complete. But the dump takes a very long time because of the "All pages with complete page edit history" dumps. Take cawiki, for instance: the dump started on February 12 and the "All pages, current versions only" was done at 0:45 the next day. The whole process, though, didn't end until February 14, 05:21. Check Wikipedia ran on the 16. This delay means the fixes made during those four days will get flagged again. Could you run your process when the "current version" dump is done, without waiting for the historic one?

--Joutbis (talk) 11:11, 22 February 2015 (UTC)

Joutbis The dump files are not available at WMFLabs until the entire dump gets finished. I have complained about this to various people and been told to stop complaining. This is assuming that dumps are actually being made (they have been very spotty recently) and WMFLabs dump directory is actually working (past year only 50% of the time). I no longer get any responses from anybody about dumps not working or directories not working. Essentially, nobody at WMFLabs or WMF cares.
I do copy enwiki's (15 day gap) and frwiki's (10 day gap) dump files, but leave the rest to when the dump actually finishes. Bgwhite (talk) 06:02, 23 February 2015 (UTC)
Joutbis Given how the dumps are managed by WMFLabs, it's not possible to run the check just after the current version dump is done. After a full scan, you can use WPCleaner to check all articles and mark as done the ones where the problem is already fixed: Bot tools, select all the error numbers below 500 and click on "Mark errors already fixed". It's not very fast, but it's fully automatic and won't do any modification on Wikipedia, I use it from time to time on frwiki. --NicoV (Talk on frwiki) 10:13, 26 February 2015 (UTC)

OK, thanks --Joutbis (talk) 19:50, 28 February 2015 (UTC)

New error: 102 - Check PMID

  Resolved

@Magioladitis, NicoV, Meno25, Josve05a, Matěj Suchánek, and ToSter: Error 102 will check for proper syntax of PMID. This is similar to error #69. See WP:PMID for more info. It "should" start checking in the next run. Bgwhite (talk) 20:26, 2 February 2015 (UTC)

Added to WPCleaner in the next release. --NicoV (Talk on frwiki) 18:24, 23 February 2015 (UTC)

Error #2 additions

  Resolved

@Bgwhite and NicoV: CHECKWIKI now detects cases such as those described in rev 10811 (i.e. <br clear=both /> and similar). -- Magioladitis (talk) 12:33, 1 February 2015 (UTC)

In Checkwiki, I just use /<\s*br\s*clear/ to find cases
WPC now detects them also, and suggests a replacement after some configuration (see example on frwiki configuration). I've only configured frwiki, other projects should do the configuration depending on the replacement they want to use. --NicoV (Talk on frwiki) 17:31, 15 February 2015 (UTC)
For enwiki, the configuration should probably be:
error_002_clear_all_enwiki={{Clear}} END
error_002_clear_left_enwiki={{Clear|left}} END
error_002_clear_right_enwiki={{Clear|right}} END

Error 61

Error number 61 (reference before punctuation)also appears when an en dash ( - ) appears instead of an em dash ( — ) in references to page numbers in periodicals and journals. This error doesn't seem to be a reference before punctuation error, so error number 61 doesn't seem to apply. Is this a bug in the wiki coding or is it really intended to exclude the ordinary dash? I have learned that the ordinary dash ("en dash") is easy (next to the number nine on the keyboard), but the "em dash" requires the typist to hold the ALT key while typing 0151. I would like to suggest a disambiguation for error number 61 to correct this.50.159.6.134 (talk) 18:47, 9 May 2015 (UTC)

Error #70 ISBN with wrong length for doi???

  Resolved

The article "Gutai group" is causing an error #70 for "10679847" in this citation (10. in the "Notes" section):

Tiampo, Ming (2013). "Gutai Chain: The Collective Spirit of Individualism" 21 (2). Positions. pp. 383–415. doi:10.1215/10679847-2018292

The doi value is good. See here The doi resolves properly. See here

It's the first citation I've encountered with a doi value, so my initial question is...why does the Checkwiki software bother with this number if it doesn't begin with "ISBN"? Knife-in-the-drawer (talk) 12:37, 11 May 2015 (UTC)

Knife-in-the-drawer I think I fixed the problem. -- Magioladitis (talk) 07:31, 12 May 2015 (UTC)

Magioladitis Yes, thank you. I added a template to the top of the article. Knife-in-the-drawer (talk) 07:47, 12 May 2015 (UTC)

Punctuation before citation

  Resolved

Per WP:REFPUNC, punctuation comes before citation. This is a long standing "house style". Furthermore there are bots such as User:BG19bot that standardize using this style. Boghog (talk) 04:44, 15 May 2015 (UTC)

Boghog I'm confused. CheckWiki only checks for this. BG19bot is my bot. Could you be more specific on this? Bgwhite (talk) 05:11, 15 May 2015 (UTC)
@Bgwhite: Never mind, I am the one that is confused. I saw an WPCleaner edit summary that stated "Reference before punctuation". I read this to mean this is how it should be rather than the error that is correcting. Sorry for the confusion. Boghog (talk) 05:17, 15 May 2015 (UTC)
Boghog I understand. You chemist are easily confused. The smart ones go into a physics discipline.  :) Bgwhite (talk) 05:29, 15 May 2015 (UTC)

Error n° 95 false positive

  Resolved

Robert Clark Young has a user link that may be a false positive. Jerodlycett (talk) 20:59, 31 March 2015 (UTC)

Jerodlycett, WordSeventeen delinked it. -- Magioladitis (talk) 16:32, 11 April 2015 (UTC)

error 6/37 on dawiki

Character '&' is allowed after the danish rules. see. And the rules says how to handle '&' in sortkeys. --Steenth (talk) 07:46, 7 May 2015 (UTC)

Steenth. The ref you gave doesn't say & is allowed. Section 1.4 says to order them based on their Danish meaning.... + means plus. Main reason for this is Stenth & sons would be sorted different that Stenth and sons. Also note the Katalogiseringsregler og bibliografisk standard for danske biblioteker is based upon the AACR2
Could you show me where on Danish Wikipedia gives the rules for sorting? Bgwhite (talk) 22:01, 7 May 2015 (UTC)

Steenth. ping. -- Magioladitis (talk) 07:30, 8 May 2015 (UTC)

Error 3 on elwiki

  Resolved

I think WPCleaner catches the list found at el:Βικιπαίδεια:WikiProject_Check_Wikipedia/Μετάφραση while CHECKWIKI script does not. -- Magioladitis (talk) 16:48, 27 June 2014 (UTC)

I think the problem is only with the last line. Now that I updated the code, I noticed that all errors shown are connected to the last line. -- Magioladitis (talk) 05:26, 8 July 2014 (UTC)

Bgwhite. Nope. The problem still occurs. -- Magioladitis (talk) 08:26, 24 January 2015 (UTC)

We are good now. -- Magioladitis (talk) 07:39, 8 May 2015 (UTC)

Error #84

  Resolved

Remember to exclude headers consisted by a sole letter from checking. -- Magioladitis (talk) 08:41, 14 January 2015 (UTC)

Magioladitis Code already does this. Wondering if another section header, say == 0-9 ==, was flagged in the article. The last dump did not flag any one character section header. Bgwhite (talk) 21:41, 22 January 2015 (UTC)

Bgwhite I can't remember anymore which page caused the problem. -- Magioladitis (talk) 09:11, 24 January 2015 (UTC)

Instances of 'subst:' in articles

  Done

do we have a scan for 'subst:' in articles? for example, for cleaning up this type of issue? Frietjes (talk) 20:59, 28 January 2015 (UTC)

@Frietjes, Magioladitis, and NicoV: A scan of the latest dump for 'subst:' can be found at User:Bgwhite/Sandbox. The listing will contain examples where 'subst:' is inside comment tags. I've added 'subst:' to the checkwiki program and results will show up under error #34.
Currently #34 scans for:
  • {{{
  • #if:
  • #ifeq:
  • #switch:
  • #ifexist:
  • {{fullpagename}}
  • {{sitename}}
  • {{namespace}}
  • {{basepagename}}
  • {{pagename}}
  • {{pagesize}}
  • {{protectionlevel}}
  • {{subpagename}}
  • {{subst:
Bgwhite (talk) 22:43, 28 January 2015 (UTC)
I've updated WPC so that it detects subst: and safesubst: as error #34. WPC detects a lot of functions as #34 (see variable functionMagicWords around line 247), but I don't think they should be added to the CW detection. --NicoV (Talk on frwiki) 11:08, 29 January 2015 (UTC)
now fixed all the ones that should be fixed in User:Bgwhite/Sandbox, so feel free to rescan, or overwrite the list with something else. Frietjes (talk) 16:40, 30 January 2015 (UTC)
@Magioladitis and Bgwhite: It seems that it also detects NUMBEROFARTICLES (Encyclopédie is detected on frwiki). Is it useful? There's nothing much we can do about this value, since it's a dynamic one. Same would apply to PAGESIZE and PROTECTIONLEVEL. --NicoV (Talk on frwiki) 08:44, 2 February 2015 (UTC)
NicoV NUMBEROFARTICLES was added for a couple of days and then removed. Magioladitis will have to answer the other question. Bgwhite (talk) 09:41, 2 February 2015 (UTC)
@NicoV and Bgwhite: let's remove both PAGESIZE and PROTECTIONLEVEL. -- Magioladitis (talk) 09:59, 2 February 2015 (UTC)

New error: unneeded headline

  Not done

Hi! I was wondering whether a headline as the first stuff of an article (excluding templates) is an error which could be also checked by CheckWiki. See example (cswiki). Matěj Suchánek (talk) 18:38, 5 February 2015 (UTC)

Matěj Unfortunately, it would create too many false positives. There are many articles that do not have a lede. Articles with tables is the most common culprits. In those cases, a headline can't be removed and the article would be a false positive. An example is List of Liberty Bowl broadcasters. Bgwhite (talk) 07:16, 6 February 2015 (UTC)
I also think there would be too many false positives. Maybe only if the headline is the same as the article title (to catch mistakes made by newbee who adds a title) ? --NicoV (Talk on frwiki) 20:06, 6 February 2015 (UTC)
I support at least Nico's idea. Matěj Suchánek (talk) 09:43, 7 February 2015 (UTC)
English dump is about ready to be processed. I'll add something to checkwiki to check to see if the first headline is the same as the article's title. After looking at the results, will decided what to do. Bgwhite (talk) 10:20, 7 February 2015 (UTC)
NicoV Matěj Magioladitis. The following articles were found in the first 1% of the dump that had the first headline the same as the article's title
IMO, La Espero, List of Pokémon, Philosophy of education, Quantum information, Religion and mythology and Hake.
I'm not to sure about this. IMO looks to be a valid case for removal, but the rest don't. AWB will remove the first section headings in all these articles. Bgwhite (talk) 20:53, 11 February 2015 (UTC)
Late to the conversation, but a thought, what if the rule was first header same as article title AND no lead. Jerodlycett (talk) 01:39, 2 April 2015 (UTC)

AWB already fixes things like these. -- Magioladitis (talk) 10:37, 7 February 2015 (UTC)

We better avoid adding this error. -- Magioladitis (talk) 21:44, 11 February 2015 (UTC)

Many false positives. Big troubles. -- Magioladitis (talk) 11:25, 22 February 2015 (UTC)

Unnecessary pipe template

  Resolved

could be interesting to find these, basically, a {{!}} inside of [[ ]]. Frietjes (talk) 17:59, 11 February 2015 (UTC)

Frietjes Magioladitis A database scan can be found at User:Bgwhite/Sandbox. Magioladitis could do an AWB bot run. F&R could be something like \[\[(.*)\{\{\!\}\}(.*)\]\] -> [[$1|$2]] This did catch some false positives, so the F&R needs to be refined. See "Weird Al" Yankovic discography.
What does the rest of the motley crew think about adding this? @Matěj Suchánek, NicoV, Meno25, Josve05a, and ToSter: Bgwhite (talk) 22:43, 11 February 2015 (UTC)

I agree. -- Magioladitis (talk) 22:53, 11 February 2015 (UTC)

Or we could fix the "Cite" thing in the toolbar, that keeps adding it everywhere, except on enwp....
  • enwp = <ref>{{cite web|title=foo|url=foo|website=[[Foo|test]]}}</ref>
  • svwp = <ref>{{webbref|titel=foo|url=foo|websida=[[Foo{{!}}test]]}}</ref>

(tJosve05a (c) 06:20, 12 February 2015 (UTC)

I also agree. Thank you for notifying me. --Meno25 (talk) 07:42, 12 February 2015 (UTC)
Yes, another nice case how to clean Wikipedia. Matěj Suchánek (talk) 14:34, 12 February 2015 (UTC)
Bgwhite and Magioladitis, the replacement should be \[\[([^\[\]]*)\{\{!\}\}([^\[\]]*)\]\] -> [[$1|$2]] which wouldn't do anything in your false-positive case. Frietjes (talk) 15:04, 12 February 2015 (UTC)
Magioladitis I've updated lists at User:Bgwhite/Sandbox. It was generated using Frietjes' regex. Bgwhite (talk) 21:26, 12 February 2015 (UTC)

@Bgwhite and Frietjes: I ran the bot with the regex and I am done. -- Magioladitis (talk) 13:26, 13 February 2015 (UTC)

@Bgwhite and Frietjes: Found exception at Toreador Song. -- Magioladitis (talk) 07:57, 15 February 2015 (UTC)

That would be an exception. However, in that article, the repeat symbol should be removed. A musical note shouldn't be there. Bgwhite (talk) 09:02, 15 February 2015 (UTC)
could update the regexp to \[\[([^\[\]\|]*)\{\{!\}\}([^\[\]]*)\]\] -> [[$1|$2]] which would exclude that case (basically no actual pipe before the escaped pipe). Frietjes (talk) 14:57, 15 February 2015 (UTC)


@Matěj Suchánek, NicoV, Meno25, Josve05a, ToSter, and Frietjes: It has been added as error #102. It should show up today in the daily scans. There is a problem with dumps in which no dumps are being produced for alot of languages. Many languages haven't been dumped since December. Bgwhite (talk) 01:03, 20 February 2015 (UTC)

rev 10835 now in AWB general fixes. -- Magioladitis (talk) 10:03, 20 February 2015 (UTC)

Now this is error #103. -- Magioladitis (talk) 11:25, 22 February 2015 (UTC)

@Bgwhite and Magioladitis:, I think this should be restricted to the target of the link, it should not detect articles where {{!}} is used in the displayed text of the link. For example, fr:Idéal de l'anneau des entiers d'un corps quadratique is detected for [[Valeur absolue|{{!}}''f''{{!}}]] where the goal is to display pipes around "f" (absolute value of "f" in maths). --NicoV (Talk on frwiki) 14:37, 23 February 2015 (UTC)

Added to WPCleaner. --NicoV (Talk on frwiki) 18:00, 26 February 2015 (UTC)

Bug in the "all priorities" page on the interface

  Resolved

I've found another bug in the "all priorities" page on the interface: basically it switches middle with low priorities. Bye! --Vittorioo (talk) 20:51, 25 April 2015 (UTC)

Vittorioo Please provide us an exact link of where this happens. Thanks, Magioladitis (talk) 07:33, 8 May 2015 (UTC)
here for example. All the "middle" priorities are "low", and all the "low" are "middle". --Vittorioo (talk) 09:21, 8 May 2015 (UTC)
Magioladitis I have submitted a pull request with a very simple fix; can you please check if it'll be enough? Thanks. --Vittorioo (talk) 09:14, 9 May 2015 (UTC)
Vittorioo Done. Bgwhite (talk) 18:05, 27 May 2015 (UTC)
Thanks! --Vittorioo (talk) 19:21, 27 May 2015 (UTC)

Dumps have stopped working.

  Resolved Due to a bug in Wikimedia's backup system, dumps have been offline for a week. WMF is working on the problem. Bug report is at T98585. Bgwhite (talk) 23:08, 9 May 2015 (UTC)

Problem fixed. -- Magioladitis (talk) 16:24, 1 August 2015 (UTC)

Old names of errors in the interface

  Resolved

Projects without a translation page, or with a partial translation, get some old names for errors in the web interface, like "No bold title" for error #1. Also missing errors from 100 to 103. Example: hrwiki. --Vittorioo (talk) 20:49, 26 May 2015 (UTC)

Vittorioo Should be fixed. Bgwhite (talk) 00:09, 27 May 2015 (UTC)
Much better, but error 31 is still wrong; still missing escape characters for tags, like in errors 38-41, 26, 101. Also it's now impossible to sort by Description. Thanks for your work. --Vittorioo (talk) 19:21, 27 May 2015 (UTC)
Vittorioo I've got free time because I'm sick. I'd like to say it is affecting my mental capabilities, but I've had none for awhile now. Added the others and now working on sorting. Bgwhite (talk) 21:32, 27 May 2015 (UTC)
Sorting should work now too. Bgwhite (talk) 21:49, 27 May 2015 (UTC)

EAN template trashes text

  Resolved

I fixed several broken ISBNs in the article "Manlio Sgalambro" using Template:EAN (i.e. they were EANs, not ISBNs). As you can see, the article got trashed.

I fiddled with the EAN template in my sandbox. If I enter

{{EAN|724383289820}}some text

then no problems--the output renders properly.

However, when I add an asterisk for a bullet item:

* {{EAN|724383289820}}some text

then everything after the EAN template gets thrown down to the next line.

I don't see a problem with how the article is formatted, and I don't see a problem with my usage of the EAN template. Therefore, I'm guessing that the problem lies either in the EAN template or in the bullet or both in combination (because I've used the EAN template previously without problems). Any ideas? Knife-in-the-drawer (talk) 15:51, 7 June 2015 (UTC)

Frietjes may know why. -- Magioladitis (talk) 21:24, 7 June 2015 (UTC)

Knife-in-the-drawer, I may have fixed it in Template:EAN/sandbox. If you use {{EAN/sandbox}}, does it work as expected for you? Does it break anything? – Jonesey95 (talk) 01:26, 8 June 2015 (UTC)
@Jonesey95: Works fine now. Thanks Knife-in-the-drawer (talk) 01:38, 8 June 2015 (UTC)
  Fixed. I have updated the main template. The problem was too much white space in the template, which was causing line breaks to appear in places where they were not desired. – Jonesey95 (talk) 01:45, 8 June 2015 (UTC)

Tool down?

  Resolved

For a couple of days, at least, I haven't been able to get to the pages that list the errors. Is wmflabs down, or has the location changed? The pages I've tried to get to are here for English and here for Simple English. Thanks. --Auntof6 (talk) 03:29, 19 June 2015 (UTC)

Auntof6 wmflabs is down. Details here. -- Magioladitis (talk) 05:14, 19 June 2015 (UTC)
Thanks for the info. --Auntof6 (talk) 06:28, 19 June 2015 (UTC)

#64

  Done The script should be become first letter case insensitive so that we spot things like these: [1], [2], [3]. -- Magioladitis (talk) 06:19, 30 June 2015 (UTC)

Magioladitis It already spots these cases. Bgwhite (talk) 18:22, 30 June 2015 (UTC)
Bgwhite how come they did not appear in the June dump then? -- Magioladitis (talk) 19:07, 30 June 2015 (UTC)
[4]. This was there since January 2015 at least. -- Magioladitis (talk) 19:59, 30 June 2015 (UTC)

Old Dump

  Resolved

I just noticed that the last dump listed is from May 15. If that's correct, there's a major bug, if it's wrong, a minor bug. Given we are down to 0 errors on en, I'm leaning towards the first.

Jerodlycett why is this a bug? -- Magioladitis (talk) 13:53, 10 July 2015 (UTC)
Jerodlycett Errors are updated every day on enwiki, plus a few other languages. If you look in the "done" column, you will see all the errors that were fixed in the past 18 hours. If you look at the main page, you will see enwiki was updated today.
The dumps for the past four months are having trouble. They are either not being run or severely delayed. The Wikipedia tech people have been messing around with things. Currently no dumps have been completed in over 20 days. Bgwhite (talk) 17:18, 10 July 2015 (UTC)

off

  Resolved

Why are "Category is english", "Reference duplication" and "Heading double" not working? --אריה ה. (talk) 08:36, 11 July 2015 (UTC)

אריה ה. because the first one replaces the word "Category" with the corresponing word in the local language, Which in the English Wikipedia is again "category". The second one does not have consensus in English Wikipedia and the third one has many false positives. Every project can choose which errors to activate. -- Magioladitis (talk) 09:04, 11 July 2015 (UTC)

403

  Resolved

Can anyone give me a with as to why this simple WebClient call (C#) failes with 403 Forbidden?

string s = webcl.DownloadString(url);

The same url works from curl and from Chrome: (Same for http, https)

https://tools.wmflabs.org/checkwiki/cgi-bin/checkwiki.cgi?project=hewiki&view=bots&id=48

Thanks 08:20, 12 July 2015 (UTC) — Preceding unsigned comment added by Kotz (talkcontribs)

Needless to say, same code works with other URLs. Is there some filtering on User Agents or so forth? Kotz (talk) 09:01, 12 July 2015 (UTC)
Solution: [5] Kotz (talk) 13:56, 12 July 2015 (UTC)
Kotz The link right above this times out for me. Users of ToolLab have minimal control of the webserver. You will need to send an email to labs-l@lists.wikimedia.org for questions. WPCleaner (Java) and AWB (.Net) both retrieve data from CheckWiki via the webserver. Bgwhite (talk) 19:18, 13 July 2015 (UTC)

ID 95

  Done

Moin Moin @ all, Moin Bgwhite, since last week the ID 95 gives many items, who have no errors. Look: Example 1 and Example 2. Yesterday I set all errors as done, today I have over 11.000 errors again. Could you have a look at. Thanks --Crazy1880 (talk) 06:54, 13 July 2015 (UTC)

Same on frwiki, where #95 list has reached 45.000 pages... The notice is not helpful to find what triggers the detection. Is it related to the addition of the Draft namespace to #95 ? This namespace doesn't exist on frwiki. --NicoV (Talk on frwiki) 10:11, 13 July 2015 (UTC)
28 pages only for arwiki. Only one false positive: ar:عبد القادر العاني. --Meno25 (talk) 13:40, 13 July 2015 (UTC)
The changes of the API Gerrit 216048 or Gerrit 160223? Thats all I know the last weeks. Regards --Crazy1880 (talk) 17:38, 13 July 2015 (UTC)
It's probably related to adding Draft space. Checkwiki is making an API call to get what Draft space is named. It's probably returning something weird and causing all the false positives. Adding draft space was essentially copy/paste of 4 lines that checked for user space, plus a modification of an if statement. Will check into this. Bgwhite (talk) 19:12, 13 July 2015 (UTC)
@Crazy1880, NicoV, and Meno25: Not all Wikipedias have draft space. Joy. Made a fix and there shouldn't be all the false positives anymore. Bgwhite (talk) 20:02, 13 July 2015 (UTC)
 Y It should be working. I will check it again tomorrow, then we see more. Thank you --Crazy1880 (talk) 07:48, 14 July 2015 (UTC)
It works, thanks. King Regards --Crazy1880 (talk) 17:34, 15 July 2015 (UTC)

  Done

The link "comments and bugs" at the bottom of this page takes you to this non-existent page which is incorrect. It should lead you here. --Meno25 (talk) 16:07, 19 July 2015 (UTC)

@Meno25: Coming here from Bgwhite's talk page. The first %26 represents a & (ampersand) but it needs to be a ? (question mark), which gives this. Percent-encoding of three other characters also breaks the URL in some browsers, so the second %26 should be a & (ampersand), and both %3D should be = (equals), i.e. https://en.wikipedia.org/wiki/Wikipedia_talk:WikiProject_Check_Wikipedia?action=edit&section=new --Redrose64 (talk) 08:31, 31 July 2015 (UTC)
Meno25 Fixed. Bgwhite (talk) 21:16, 31 July 2015 (UTC)
@Redrose64: Thank you, sir, for your comment. Bgwhite Thank you. However, the link is still not working for me. I will assume that you haven't uploaded the corrected code to the server. --Meno25 (talk) 06:34, 1 August 2015 (UTC)
Meno25 It's working for me. Make sure it says at the bottom, "Version 2015-07-30", as that is the updated code. If it says that and still has problems, give a yell. Bgwhite (talk) 08:18, 1 August 2015 (UTC)
Bgwhite Confirmed working. Thank you very much. --Meno25 (talk) 08:37, 1 August 2015 (UTC)

Error #48

  Resolved

The proper way to resolve Error #48, a link to the article title in the text, is to replace the double brackets with triple quotes rather than deleting the entire link, e.g., [[article title]] → '''article title'''. This leaves the article appearance unchanged. Yours aye,  Buaidh  20:49, 25 July 2015 (UTC)

Not always, and often no, because bold shouldn't be used everywhere. It does change the appearance, but it's often better to remove the bold. Cf. MOS:BOLD. --NicoV (Talk on frwiki) 21:02, 25 July 2015 (UTC)

A troll or a bug?

  Resolved

Hello. I found this project a couple of weeks ago. It is very helpful, thank you very much. I already cleaned all 4,500 the high priority errors in our wiki. But there is a strange problem. Every day I can see hundreds of errors that are marked as done, not by me, and without any actual article fix. And I checked: the error is indeed there. Can it be a bug that marks an article as done when I fixed another error in it? And if can't, and this is indeed a troll, can I somehow restore these marks? Thank you, IKhitron (talk) 13:43, 26 July 2015 (UTC)

IKhitron what is your wiki? -- Magioladitis (talk) 14:30, 26 July 2015 (UTC)
It's hewiki, Magioladitis. Thanks, IKhitron (talk) 14:36, 26 July 2015 (UTC)
IKhitron I don't recall any code that marks an error as done on the checkwiki end. WP:WPCleaner does, but that would require human intervention. A new dumpfile just came out, so all the errors were just updated. I have a feeling somebody is marking "done" to errors that they feel are not errors. Bgwhite (talk) 20:47, 27 July 2015 (UTC)
I see. The problem continues on the new dump information too. Is there a posiibility to unmark done or to observe done errors at least week ago? Thank you, IKhitron (talk) 08:57, 28 July 2015 (UTC)
IKhitron There is a log file. I can give you a listing of articles by error number. Which error do you want first? FYI... logs says there were 43,302 errors. Bgwhite (talk) 20:55, 28 July 2015 (UTC)
Thank you very much, Bgwhite, but I don't think it will help. First of all, if the log has just article name but not context, it takes a lot of time usually to find the exact error place. Second, I can't ask you every day show me dozens of files. I believe we can't do nothing, if there is no ability to see somehow all the dump check results ignoring any dones. And if there were more than 43,000 errors, and now we have about 31,000, and I fixed 4,500 from the last dump, it means we lost 7,500 errors. Oh, cat! IKhitron (talk) 11:29, 29 July 2015 (UTC)
IKhitron Log files show context. I'd normally say another dump would be done in two weeks, but WMF employees are screwing up on producing dump files. They are doing a new way to be more "efficient", which currently means less dumps and who knows when they next one will be. Bgwhite (talk) 18:26, 29 July 2015 (UTC)
If it shows context, Bgwhite, is there any possibility to see all of it in our wikipedy? Thank you, IKhitron (talk) 20:45, 29 July 2015 (UTC)
Bgwhite? Please. IKhitron (talk) 11:13, 6 August 2015 (UTC)
IKhitron Sorry, I completely forgot. Which errors do you want? The log file is big and hard to understand. I can break it down into errors that would be easier. Bgwhite (talk) 18:52, 6 August 2015 (UTC)
This is exactly the problem, Bgwhite, I don't want you to spent a lot of your time on my problems. It's just not fair. If you can give me all, so I'll try to find all I need - great. If I'll fail - my fault. If you can't - I'll survive without it. If you don't sure you can give me the information, you can check - I have the sysop rights. If it's not enough, and you need enwiki sysop to recommend me - I can arrange that. Thank you again. IKhitron (talk) 20:08, 6 August 2015 (UTC)
IKhitron I sent an email on the link for the file. I have no problems giving the info. I'm not privy to any "private info" and all I'm giving is public. Ewww, sysops are evil, especially the enwiki ones :) Bgwhite (talk) 21:34, 6 August 2015 (UTC)
[6] Thanks a lot again for your help, Bgwhite. IKhitron (talk) 21:59, 6 August 2015 (UTC)
N3, Bgwhite, I'm no so stupid as you believe. It took me less than 5 minutes in Excel to create a normal list of each error so I could work... IKhitron (talk) 14:18, 7 August 2015 (UTC)

False positives in ISBNs

  Resolved

Hello, the script is flagging four entries with valid ISBNs here:

  • ISBN 9991390634 is a real book
  • Same with ISBN 99913-90-11-1, see here. 99913 is the ISBN prefix for Andorra.
  • As for ISBN 9998736102, Amazon accepts it as a real ISBN
  • I haven't found external references for ISBN 9999532357, but the checksum is correct.

Is it possible that ISBNs starting with 999 get flagged regardless of checksum? --Joutbis (talk) 17:58, 27 July 2015 (UTC)

Joutbis I think it's because they are in reserved ranges. For Andorra, the ranges are listed below. As you can see, range 6050000-9999999 has a length of 0, and I think it means that the range is currently reserved (at least, that's what #72 is doing).
--NicoV (Talk on frwiki) 18:25, 27 July 2015 (UTC)
Joutbis, if you try the ISBN converter of the Library of Congress (check Hyphenate ISBNs), you'll see that you get an error message when trying to convert the first ISBN, 9991390634, which seems to confirm my analysis. And here, you have <Length> - [length] defined for range (zero indicates range not defined for use) --NicoV (Talk on frwiki) 20:21, 27 July 2015 (UTC)
    <Group>
      <Prefix>978-99913</Prefix>
      <Agency>Andorra</Agency>
      <Rules>
        <Rule>
          <Range>0000000-2999999</Range>
          <Length>1</Length>
        </Rule>
        <Rule>
          <Range>3000000-3599999</Range>
          <Length>2</Length>
        </Rule>
        <Rule>
          <Range>3600000-5999999</Range>
          <Length>0</Length>
        </Rule>
        <Rule>
          <Range>6000000-6049999</Range>
          <Length>3</Length>
        </Rule>
        <Rule>
          <Range>6050000-9999999</Range>
          <Length>0</Length>
        </Rule>
      </Rules>
    </Group>

Joutbis Ran into the same problem on enwiki with that book. We ended up just using OCLC 801960401 Bgwhite (talk) 20:31, 27 July 2015 (UTC)

Error 69 updated

  Resolved

@NicoV: After a discussion I had with Bgwhite, he changed

elsif ( $test_text =~ / \[\[ISBN\]\]\s*([:]|[-]|[#])+\s*\d/g ) {

with

elsif ( $test_text =~ / \[\[ISBN\]\]\s*([:]|[-]|[#])*\s*\d/g ) {

Just letting you know in case you have to change something in WPCleaner. -- Magioladitis (talk) 08:09, 1 August 2015 (UTC)

I haven't tested the change yet, so it is not live. Bgwhite (talk) 08:14, 1 August 2015 (UTC)
Thanks for the notice. I will check but I think either I already detect them or I fail to detect all [[ISBN]] stuff. --NicoV (Talk on frwiki) 13:58, 3 August 2015 (UTC)

Non-breaking spaces detected by #16

Hi Bamyers99. It seems that recently (maybe a few weeks), #16 started to detect also non-breaking spaces, but nothing in the description mentions it. Is it normal? On frwiki, we now have more than 300k articles in the list for #16 due to this...

When I run the dump analysis on frwiki with WPCleaner (twice a month), at the end I perform a cleanup on the CheckWiki lists:

  • For each article in a list, WPCleaner checks if it also detects the error
  • If it doesn't detect the error (which is the case for nbsp with #16) it then asks checkarticle.cgi to confirm if the error is still present
  • If checkarticle.cgi says that the error is not present anymore, it then asks CW to mark the page as fixed

With this list of 300k, it takes a very long time to process... What should I do? Also detect nbsp for #16 or wait for a modification of CW? --NicoV (Talk on frwiki) 18:05, 6 March 2019 (UTC)

@NicoV: Per ID16 in german above, #16 was enabled for all non-enwiki wikis on 13 February 2019. I have removed non-breaking space from the Unicode control character check. --Bamyers99 (talk) 18:50, 6 March 2019 (UTC)

False positive on .NET Core for sentences starting with ".NET" that follow a sentence with <ref>s

I just undid this change on .NET Core:

https://en.wikipedia.org/w/index.php?title=.NET_Core&diff=886087039&oldid=885975798

The tool being used evidently has a false positive for sentences that start with a period, as is sometimes the case in articles discussing .NET, if they follow a sentence that has <ref>s at the end. It erroneously thinks these instances are "Reference before punctuation" violations, and removes the space between sentences and the period at the beginning of ".NET". --Dan Harkless (talk) 03:43, 6 March 2019 (UTC)

Thanks for the report Dan Harkless. The modification is done manually by the editor, so I'm ping Vieque who did this edit. I don't know if the page was reported by CheckWiki or not, I will let this part to the people maintaining CheckWiki. But I will check if I can handle this situation in WPCleaner to avoid reporting this as an error and suggestion a replacement. --NicoV (Talk on frwiki) 09:53, 6 March 2019 (UTC)
Danke schön. The checking tool should probably ignore cases where a period is immediately followed by some non-whitespace character besides '<'. If the few false negatives that would be let through when editors fail to put spaces after their punctuation were considered too worrisome, it should at least ignore cases like:
Text.<ref>ref_stuff</ref> .NET is dot-nifty.
where there's already a period before the preceding <ref>(s). --Dan Harkless (talk) 10:12, 6 March 2019 (UTC)
About CheckWiki, it seems that it really detects this as a #61 error, it's not only WPCleaner which detects it. --NicoV (Talk on frwiki) 18:17, 6 March 2019 (UTC)
@Dan Harkless and NicoV: I have changed CheckWiki to only report #61 if there is a whitespace character after the punctuation. --Bamyers99 (talk) 18:45, 6 March 2019 (UTC)
Cool, thanks. Although as I mentioned, you should probably also report it when there's a '<' after the punctuation, so you catch cases like:
Text<ref>ref2</ref>,<ref>ref1</ref>
where the original ref(s) were placed correctly, but then someone comes along and adds an additional ref before the punctuation. --Dan Harkless (talk) 22:16, 6 March 2019 (UTC)
I've modified WPCleaner to avoid detection for constructions similar to .NET. --NicoV (Talk on frwiki) 14:15, 9 March 2019 (UTC)

Regenerating lists

Hi. I just finished a trial for my bot, which is intended to cut down on the backlog of #17 errors. But, as far as I know AWB can't automatically mark pages as fixed, and I don't want to click each one manually - is there an option to have the project rebuild the list at the end of the bot run, so that it automatically accounts for errors solved? --DannyS712 (talk) 21:24, 10 March 2019 (UTC)

alternatively, I can use AWB to export a list of the pages edited - could that be used to remove them from the list? --DannyS712 (talk) 21:26, 10 March 2019 (UTC)
@DannyS712: For enwiki and a couple of other large wikis, modified articles are re-scanned at the end of the day. This will remove the articles from the list. --Bamyers99 (talk) 22:27, 10 March 2019 (UTC)
@Bamyers99: So if I don't mark pages as being done, thats okay? --DannyS712 (talk) 22:31, 10 March 2019 (UTC)
@DannyS712: Not marking as done is okay. They will automatically get marked as done at the end of the day. Not marking happens all of the time when someone who knows nothing about the Checkwiki project fixes an error when they are editing the page for some other reason. --Bamyers99 (talk) 22:41, 10 March 2019 (UTC)

Improvements for #16

Hi, the current regex for #16 contains several characters that are not converted to something readable in the list: the list of conversions lacks \x{202B} and \x{200F} compared to the regular expression. --NicoV (Talk on frwiki) 09:11, 11 March 2019 (UTC)

@NicoV: Done.

CheckWiki pages unaccessible

Hello. For several hours, the CheckWiki pages are unaccessible, at least for cs-wiki (see f. e. https://tools.wmflabs.org/checkwiki/cgi-bin/checkwiki.cgi?project=cswiki&view=high). The page remains blank in the "read" state, and after bunch of time finally displays 504 Gateway Time-out. See also cs:Pod lípou (technika)#CheckWiki nefunguje?. --Vachovec1 (talk) 19:32, 11 March 2019 (UTC)

@Vachovec1: I restarted the web server. --Bamyers99 (talk) 19:53, 11 March 2019 (UTC)
It's working now, thanks! --Vachovec1 (talk) 20:00, 11 March 2019 (UTC)

False positives for #46

Hello. Looks like #46 (Square brackets without correct beginning) generates false positives when there is a wikilink inside the image description.

See:

--Vachovec1 (talk) 11:07, 12 March 2019 (UTC)

@Vachovec1: I have just fixed a bug in #46 that caused a false positive on the wrong square bracket pair. The real #46 is further down in the article. cs:Mimoúrovňová křižovatka has a real #46 further down as indicated by check article. --Bamyers99 (talk) 18:50, 12 March 2019 (UTC)
Ah, yes. It definitely looks like the wrong square bracket pair was/is reported. The listed articles have a real #46 error, but it's further down in the article. --Vachovec1 (talk) 20:53, 12 March 2019 (UTC)

In addition to already covered ones, some other cases could be checked for. In particular, missing / invalid number of slashes/colons. Matěj Suchánek (talk) 10:04, 15 March 2019 (UTC)

Malformed non-breaking space entity

Hi, I'd like to propose checking for malformed &nbsp; entities. Eg. typos like &(nsbp|nbps|bnsp); or missing ampersands. Matěj Suchánek (talk) 09:46, 15 March 2019 (UTC)

@Matěj Suchánek: As the only active server-side software maintainer, I am only doing bug fixes / keep the project running type of work. Feel free to clone the Perl GitHub project and submit patches or I can add you as a maintainer of the Checkwiki project on Toolforge. --Bamyers99 (talk) 18:22, 15 March 2019 (UTC)
I see, thanks for doing at least that. If I were familiar with Perl scripting, I would happily do... Matěj Suchánek (talk) 09:26, 16 March 2019 (UTC)

Automatic fixing for #46

Hi, I added some automatic fixing in WPC for some cases of #46 (square brackets without correct beginning), see more details in Wikipedia:Bots/Requests for approval/WikiCleanerBot 3 where I asked to run it as a bot on enwiki (if anyone want to participate in the discussion). I already ran it on frwiki which allowed me to fix several hundreds pages. It should work on other wikis if someone is interested. --NicoV (Talk on frwiki) 12:17, 3 April 2019 (UTC)

Possible error in the "List of errors page" and the errors generated

See WT:MOS#source substantiating latter part of sentence. It's not correct that a <ref>...</ref> may not appear immediately before a punctuation mark. See that thread, and MOS:REFPUNCT. When it does happen properly, it will almost always be a bracket (usually a round one) or a dash. It would probably be better to limit this to specific punctuation marks, and adjust the wording to say so.  — SMcCandlish ¢ 😼  08:20, 4 May 2019 (UTC)

@SMcCandlish: The only characters that are considered punctuation by CheckWiki are .,?:;! --Bamyers99 (talk) 16:51, 4 May 2019 (UTC)
Let's have the error message and documentation here be clearer about that. "May not appear immediately before a punctuation mark" is confusing and incorrect, as well as apt to get people to ignore what the REFPUNCT guideline actually says.  — SMcCandlish ¢ 😼  20:51, 4 May 2019 (UTC)
I've updated the List_of_errors material, but the actual error message emitted by the checker should be more specific.  — SMcCandlish ¢ 😼  21:14, 4 May 2019 (UTC)

Cleared all of #62

Hello fellow project mates,

It took a while, but I finally cleared the entire section of 62. All 355 of them. One less category we have to worry about for now!

AmericanAir88(talk) 14:39, 31 May 2019 (UTC)

#104 and Wikipedia’s policies

User Mathglot explained, that the error definition "#104 Unbalanced quotes in ref name or illegal character." is not according tot the Wikipedia’s policies. Please check your error definitions, --GünniX (talk) 06:00, 6 June 2019 (UTC)

I don't see any problem with the edit made by GünniX. The reference name had one quotation mark, so it was indeed unbalanced. I recommend using a better edit summary that links back to this page and references the error ID that you are fixing. – Jonesey95 (talk) 11:11, 6 June 2019 (UTC)

checkarticle.cgi not responding (or being very very slow?)

Hi. It seems that checkarticle.cgi is not responding any more. Does anyone know how to check it? Example --NicoV (Talk on frwiki) 21:28, 4 August 2019 (UTC)

@NicoV: I have just now restarted the webserver. --Bamyers99 (talk) 22:04, 4 August 2019 (UTC)
@Bamyers99: Thanks a lot! I think that checkarticle.cgi isn't very optimized and may take more than 60 seconds to analyze big articles, so a timeout is happening. --NicoV (Talk on frwiki) 11:06, 9 August 2019 (UTC)

False positives for #17

Hi. When running my bot to fix #17 errors, I stumbled on several articles which are false positives : they are redirects to a category in which they are categorized, see for example Academic Journal of Animal Diseases (and all Academic Journal... in #17). In WPC, I've currently modified #17 to avoid detection in redirects: it's not a perfect solution, but it avoids a lot of false positives. --NicoV (Talk on frwiki) 11:05, 9 August 2019 (UTC)

@NicoV: I have fixed it so that the redirected to category is not counted as a page category. --Bamyers99 (talk) 01:44, 10 August 2019 (UTC)

German daily scan

Moin Moin together, Bamyers99, since yesterday the daily scan for the german CheckWiki isn't running any more. Could you please have a look? Thanks --Crazy1880 (talk) 07:21, 19 May 2019 (UTC)

@Crazy1880: Everything looks normal to me for dewiki. See HTML-Auszeichnung <small> in Referenz, sub oder sup.
Per the log file:
Start time:                    Sun May 19 00:05:13 2019
Articles checked:              1610
Errors found:                  1833
Program run time:              0 hours, 8 minutes and 24 seconds

--Bamyers99 (talk) 14:34, 19 May 2019 (UTC)

Strange Bamyers99, it looks like ID 3, for example, wasn't running for two days. Normally there where round about 20 items. I will have a look at the next days. Regards --Crazy1880 (talk) 14:37, 19 May 2019 (UTC)
Moin Moin Bamyers99, today it seems not running, but ID 63 do! Could you double-check that? Regards --Crazy1880 (talk) 04:53, 20 May 2019 (UTC)
@Crazy1880: The daily scanner was stuck and not updating. I restarted it. --Bamyers99 (talk) 18:16, 20 May 2019 (UTC)
Moin Bamyers99, many thanks. Is it possible, that I could restart this too? Regards --Crazy1880 (talk) 18:30, 20 May 2019 (UTC)
@Crazy1880: Only this tools maintainers can restart it. --Bamyers99 (talk) 18:35, 20 May 2019 (UTC)
Moin Moin Bamyers99, after the ddos-attacs the german daily scan ins't running again. Could you please restart them? Thanks a lot --Crazy1880 (talk) 18:34, 9 September 2019 (UTC)
@Crazy1880: The scanner job hung on September 7 at about 1:30 UTC. I have just now restarted it. --Bamyers99 (talk) 18:51, 9 September 2019 (UTC)

Add Strip Markers to the check?

See Help:Strip markers for the documentation. Seems like there's a few hits around.Pawngpawng (talk) 17:27, 17 October 2019 (UTC)

@Pawngpawng: Strip markers are not present in the raw wikitext. Strip markers would only be present after the raw wikitext is parsed/rendered for output to the browser. Check wiki only scans the raw wiki text. --Bamyers99 (talk) 18:51, 17 October 2019 (UTC)

Check Wikipedia down

@Bamyers99:

Check Wikipedia is completely down. Please check it and maybe restart, thank you. Doc Taxon (talk) 11:10, 24 October 2019 (UTC)

Check Wikipedia
Software error:
Could not connect to database: Can't connect to MySQL server on 'tools-db' (111 "Connection refused")
For help, please send mail to this site's webmaster, giving this error message and the time and date of the error.


Clicking on the projects on https://tools.wmflabs.org/checkwiki/ gives this:

Software error:
Could not connect to database: Can't connect to MySQL server on 'tools-db' (113 "No route to host")
For help, please send mail to this site's webmaster, giving this error message and the time and date of the error.

Doc Taxon (talk) 11:28, 24 October 2019 (UTC)

@Doc Taxon: The hardware that runs the Checkwiki database is being worked on (T227540) and service is disrupted. --Bamyers99 (talk) 15:13, 24 October 2019 (UTC)
@Doc Taxon: Everything should be up and running. Here is a more specific ticket T236384. --Bamyers99 (talk) 16:37, 24 October 2019 (UTC)
The database server is still having issues per T236420. --Bamyers99 (talk) 19:14, 24 October 2019 (UTC)
thank you, Doc Taxon (talk) 19:50, 24 October 2019 (UTC)

References: ref file C, ref file HOME, etc

Do we need to do something automatically with such links: file:///C: (1), <ref>file:///home</ref> (2)? — Preceding unsigned comment added by Fractaler (talkcontribs)

@Fractaler: Well, they're not useful references, so they need some kind of fixing, not sure what could be done automatically though. I'll add that when I've seen file paths in the past, it's often been a indicator of potential paid editing (someone handed someone else a Microsoft Word document and paid them to post it). creffpublic a creffett franchise (talk to the boss) 13:35, 19 November 2019 (UTC)
@Creffett: It would be nice to automatically prevent the addition of such links --Fractaler (talk) 13:56, 19 November 2019 (UTC)
@Fractaler: Could be a case for a filter, maybe suggest it at the edit filter requests board? I'm curious how often legitimate file paths show up in edits (maybe we could restrict it to refs and links). creffpublic a creffett franchise (talk to the boss) 18:28, 19 November 2019 (UTC)
@Creffett: Thank you for the link! --Fractaler 09:00, 20 November 2019 (UTC)

False positives in #71

There are false positives in Wikipedia:CHECKWIKI/WPC 071 dump.

Example: Hiyokoi: isbn=X78-4-08-867044-7

The string occurs within a URL: http://books.shueisha.co.jp/CGI/search/zen_list.cgi?siries_isbn=X78-4-08-867044-7&siries_kanren_isbn=&mode=2

Can the code that generates this report be modified to ignore patterns within URLs? I believe that this problem also occurs in the other ISBN reports (69, 70, 72, and Wikipedia:WikiProject Check Wikipedia/ISBN errors). I can look for other examples if that is helpful. – Jonesey95 (talk) 20:51, 3 December 2019 (UTC)

Hi Jonesey95. This is a problem with WPCleaner. Can you post a request on its talk page? --NicoV (Talk on frwiki) 10:09, 9 December 2019 (UTC)
Done. Thanks. – Jonesey95 (talk) 15:06, 9 December 2019 (UTC)

Checkwiki is down on WMFLabs

Hi. Checkwiki returns a 503 error. --NicoV (Talk on frwiki) 13:32, 22 December 2019 (UTC)

@NicoV and Doc Taxon: The webservice failed for some reason and did not automatically restart. I started it manually. --Bamyers99 (talk) 14:28, 22 December 2019 (UTC)

Apostrophes - should edits be made?

Whilst using WPcleaner recently, I've come across a few changes from ’ to ', under Spelling and typography. See Special:Diff/934429763 for an example.

If this is the only change to be made, should it be done?

I was thinking in terms of AWB users not being permitted to make insignificant or inconsequential edits under wp:AWBRULES 4. I assume the same principle applies to all tools?

Thanks, ~~ OxonAlex - talk 16:11, 6 January 2020 (UTC)

This is not a cosmetic edit. It is a MOS-related fix. Editors should manually inspect any edits that replace curly quotes with straight quotes, however. Sometimes editors incorrectly use curly quotes in place of the ʻokina, the grave accent mark, or the prime mark, and we should not replace those. We should also not replace curly quotes used deliberately, as in quotation mark. Changing curly to straight quote marks can also change formatting of text by introducing bold or italic formatting where it may not have been intended. – Jonesey95 (talk) 16:30, 6 January 2020 (UTC)

Ancient dump reported

This seems to indicate that there has been no dump analyzed in forever, but I know daily updates are done. Jerod Lycett (talk) 11:17, 6 January 2020 (UTC)

@Jerodlycett: The enwiki and dewiki semi-monthly dumps are too big to be scanned in a reasonable amount of time. enwiki, dewiki, eswiki, frwiki, arwiki, cswiki, plwiki are scanned daily via recent changes. --Bamyers99 (talk) 15:42, 6 January 2020 (UTC)
Reasonable amount of time? Even weekly scans would be good. Headbomb {t · c · p · b} 16:30, 6 January 2020 (UTC)
I don't know much about dumps, but this page appears to indicate that the previous dump for enwiki was completed on 20 December, and the current dump is still being built (as of 6 Jan). Judging from the history of CheckWiki 069 report, I'd guess that dump analysis is being performed every time a full dump is available, about twice a month. Does this help, or am I on the wrong track? – Jonesey95 (talk) 17:40, 6 January 2020 (UTC)
Dump analysis takes a long time for enwiki... For the pages I generate twice a month for enwiki (not all CW errors), it takes about 20 hours on my laptop (and laptop is not very usable during the analysis). I don't know how Tool Labs compares to my laptop, but I think it restricts computing power for each task. I don't know either how resource consuming CW analysis is compared to WPC analysis. --NicoV (Talk on frwiki) 19:02, 6 January 2020 (UTC)
And also this page shows that the dump from 1st of January is still running for some parts. And I think the dumps are available on Tool Labs only when they are 100% completed. For WPC, I only use one file that I manually download, which is usually ready after one day of dump analysis. --NicoV (Talk on frwiki) 19:05, 6 January 2020 (UTC)

False positive on Error 03

@Magioladitis: Turns out {{Schubert's compositions (references)}} has the reflist in it. This is an example of where it's being used. If you could whitelist that template it would be useful. (Per the diff, only just discovered it myself). Jerod Lycett (talk) 23:36, 31 December 2019 (UTC)

@Jerodlycett: Mags is not active on this project anymore. I have updated the reflist template whitelist. --Bamyers99 (talk) 00:23, 1 January 2020 (UTC)
@Bamyers99: Should we update the list on the main page then of participants? Jerod Lycett (talk) 00:35, 1 January 2020 (UTC)
The common practise is not to have templates with reflist inside. -- Magioladitis (talk) 19:09, 1 January 2020 (UTC)

Jerod Lycett,Bamyers99 I suggest that we remove the Reflist outside because it's easier to control reflist section when not inside the template, it allows more flexibility, it may be adjusted per page, etc. -- Magioladitis (talk) 19:12, 1 January 2020 (UTC)

I don't disagree, but I work on this stuff here. I'll bring it up on the talk page there. It's only used on 23 pages so it should be easy enough to fix them all. Jerod Lycett (talk) 19:20, 2 January 2020 (UTC)
@Magioladitis: I brought it up in the talk going on a week ago. With no response, I suggest it's open for you making the change as you had before. It would appear to only truly affect the one page with a non-template reference. Jerod Lycett (talk) 12:17, 9 January 2020 (UTC)

Bamyers99 I am semi-active to the project pending some discussions on the CHECKWIKI errors that, per ARBCOM, have to be done without me and I am still waiting two years for that. I have been banned from discussing the nature of the CHECKWIKI errors. I think the banned was supposed to be temporary until the discussion is over. But the discussion does not seem to ever occur. I can still help with technical issues and solutions. I can also perform bot edits and discuss practices we used in the past. -- Magioladitis (talk) 19:14, 1 January 2020 (UTC)

@Magioladitis: you are prohibited from discussing WP:COSMETICBOT-related issues such as whether CW Error #12 should be considered cosmetic or not. You are not prohibited from discussion of CHECKWIKI fixes in general, or prohibited from making CHECKWIKI fixes in general (even cosmetic ones), so long as they are made through your bot accounts, with a formal BRFA for a specific task. You can also, if something is unclear, ask at WP:BOTN if something is considered cosmetic or not. Headbomb {t · c · p · b} 19:33, 2 January 2020 (UTC)
@Headbomb: Yes, I know that. This is the reason I am semi-active. I am waiting for the discussion of which CHECKWIKI errors are considered cosmetic or not to conclude. There are still 30+ errors in the list with a question mark. Some of them are not AWB related. I can't contribute further to the discussion. What I can do is help by discussing common practices of how we dealt with technical issues so far. Adding or modifying the list could be considered violation so I am not contributing to that area either. -- Magioladitis (talk) 21:47, 2 January 2020 (UTC)
Or tackle any of the 100+ marked "No", such as CW Error #03. Headbomb {t · c · p · b} 21:56, 2 January 2020 (UTC)
There are two things to deal: What is to be done and how is to be done. -- Magioladitis (talk) 00:58, 4 January 2020 (UTC)
I'll check whether my bot can fix this. But I think it can't be fixed solely and there were also times it was failing to fix a page due to lack of references section. -- Magioladitis (talk) 13:28, 4 January 2020 (UTC)

Error 95 and {{In use}}

Old revision of Jerome and Jeremiah Valeska was flagged as having the signature. It was in the In use template though. I think this qualifies as a false positive even though it's an incorrect use of the template which seeks five tilde instead of four. Jerod Lycett (talk) 11:18, 9 January 2020 (UTC)

@Jerodlycett: Operating as designed. Not a bug. --Bamyers99 (talk) 17:51, 9 January 2020 (UTC)

False positive (?) on #95

Hi. I was checking Check Wikipedia and for "Editor's signature or link to User space" it says there is a signature on User talk:CyberTroopers. (see https://tools.wmflabs.org/checkwiki/cgi-bin/checkwiki.cgi?project=enwiki&view=detail&title=User%20talk:CyberTroopers). I'm not sure if its a bug or not but shouldn't talk pages have signatures? Thanks. OkayKenji (talk page) 09:57, 29 December 2019 (UTC)

@OkayKenji: Checkwiki was designed for scanning pages in the article namespace only. --Bamyers99 (talk) 15:42, 29 December 2019 (UTC)
Thanks, but if its designed for scanning namespaces only...I was wondering why it came up in the list though? (I didn't specifically scan on that page, it just came up in said list) OkayKenji (talk page) 15:56, 29 December 2019 (UTC)Nevermind, perhaps I missing something but its okay. Thanks! OkayKenji (talk page) 07:11, 31 December 2019 (UTC)
@Bamyers99 and OkayKenji: This is still in the list, OkayKenji was correct. It is the only one left in it right now even. I have no idea how it got there, but it's there. This is a bug report. Jerod Lycett (talk) 11:12, 9 January 2020 (UTC)
@Jerodlycett: I marked the page as done. I have checked the code. It only scans namespace 0. checkarticle.cgi only reports errors to the callee, it does not insert them into the database. I am not researching this any further. --Bamyers99 (talk) 17:46, 9 January 2020 (UTC)
@Bamyers99: Yeah. It's back. I am so confused as to how this one page in particular is showing up. I can't find anything even redirecting to it that could cause it. Do we have access to logs? Jerod Lycett (talk) 08:02, 15 January 2020 (UTC)
@Jerodlycett: When I marked it as done before, I forgot to mark the other 3 issues for that page as done. Checkwiki periodically rescans old issues to see if they are still an issue and it rescanned this page because the other 3 issues were still in the database. I have now marked all 4 issues as done, so hopefully it won't show up again. --Bamyers99 (talk) 23:45, 15 January 2020 (UTC)

Magic words

Could someone (e.g., with AWB) go through [https://en.wikipedia.org/w/index.php?title=Special:Search&limit=100&offset=0&ns0=1&search=-intitle%3Aelectoral+-intitle%3Aelection+insource%3A%2F__NOEDITSECTION__%2F&advancedSearch-current=%7B%7D&searchToken=dr8ah9bin9udl39puis3hqp6z these search results] and remove all of the NOEDITSECTION magic words? I looked at everything in that list, and with the exception of the Main Page itself, it shouldn't have that switch on the page. WhatamIdoing (talk) 19:10, 23 January 2020 (UTC)

  Done — JJMC89(T·C) 07:42, 24 January 2020 (UTC)
Special:diff/937311071 IKhitron (talk) 13:10, 24 January 2020 (UTC)

Error 95 and {{Under construction}}

The placedby parameter is for a user name... should this be whitelisted or should we suggest they don't do that? Jerod Lycett (talk) 11:52, 2 February 2020 (UTC)

@Jerodlycett: Fixed. --Bamyers99 (talk) 20:54, 2 February 2020 (UTC)

Leakage

Two things for you:

  1. The discussion at Wikipedia talk:Teahouse#Reply tool brings me here to ask how often, in your experience, unbalanced formatting (e.g., I make a word bold and forget to close it, or I open a span tag and forget to close it) causes problems on talk pages. If it were possible to clean up some of that automagically (e.g., automagically inserting the closing tag at the end of my comment), would that be desirable? If the devs say "everything's a tradeoff", then how much would you trade for that? (You're welcome to click the link and play with the tool in my sandbox over there. If you want to do really extensive testing, please try the Beta Cluster, preferably after the next SWAT deployment slot, when two bugs will get fixed there.)
  2. Please note mw:New requirements for user signatures. It won't do everything, but I hope this will eventually result in fewer lint errors from signatures.

(Please ping me.) Whatamidoing (WMF) (talk) 18:21, 9 March 2020 (UTC)

Hi Whatamidoing (WMF). For your first question, I think that would require testing, but I think it would be nice if this tool could close unbalanced formatting, as a reply should probably be self-contained. I'm not sure it's really causing trouble right now, it would just help reducing the number of new Linter errors... --NicoV (Talk on frwiki) 11:38, 30 March 2020 (UTC)
Thanks for the reply, NicoV. The simplest (from the Editing devs' POV) way to clean up tags is just to run the talk page through Parsoid, so at least we'd know which problems to expect. But as a longer-term thing, they'd like to extend wikitext to give us a "put this mess in a container" syntax. Keep an eye out for anything about "Multi-line comments". The WMF trying to make some risk-reduction choices until the COVID-19 situation is more stable, and new wikitext syntax is always (somewhat) risky, but so this will probably not happen soon, but I think it'll be an even better solution in the end. Whatamidoing (WMF) (talk) 16:10, 30 March 2020 (UTC)

Bot approval request

  Resolved

Hi. Regarding the new errors added to WPCleaner, I've requested approval for fixing some of them with my bot, feel free to comment on them:

--NicoV (Talk on frwiki) 16:35, 30 March 2020 (UTC)

False positives for #111 on frwiki

  Resolved

Hi. I've found a lot of false positives for #111 on frwiki, and I don't understand what could be causing it. A few examples:

  • fr:Vendôme: CW reports <ref>L'attribution de ses œuvres..., while there's 2 calls {{Références|}} later in the text (Références is listed in error_111_templates_frwiki)
  • fr:Vendrennes: CW reports <ref>{{Article|titre=Une énorme fleur..., while there's a call {{Références|}} and {{Références nombreuses| later in the text (both Références and Références nombreuses are listed in the configuration

It seems there are dozens of pages like that on the list for frwiki. Is it a bug in CW or is there something to change in the configuration? --NicoV (Talk on frwiki) 07:03, 1 April 2020 (UTC)

@NicoV: I have just fixed this. It was hard coded to check for <references and if found, it would ignore the template list. It found <references group="N" /> in the two examples above. I have fixed it to check the templates even if <references is found. Also, it uses the template list from #003, and ignores the #111 template list. I have not changed this as both template lists should be identical. --Bamyers99 (talk) 19:45, 1 April 2020 (UTC)
Thanks a lot Bamyers99! --NicoV (Talk on frwiki) 19:51, 1 April 2020 (UTC)

Svwp 

Hello! The svwp dump is usually updated with new articles twice a month. It has been 33 days since the last time. Any reason why? I was contacted on svwp since I'm a contact perosn there. Jonatan Svensson Glad (talk) 19:24, 3 April 2020 (UTC)

@Josve05a: There was only one dump for March per the Xmldatadumps-l mailing list --Bamyers99 (talk) 20:02, 3 April 2020 (UTC)
Ah, thanks! Jonatan Svensson Glad (talk) 20:34, 3 April 2020 (UTC)

New errors with WPCleaner

Hi. If anyone is interested, I've added a few errors to WPCleaner in the last weeks:

  • 544: Missing end template in a pair. It's currently disabled on enwiki. It requires configuration (see frwiki for an example) to define which templates should go by pair (begin/end)
  • 545: Template with deprecated parameter. It's current enabled on enwiki. Initial configuration has been defined for a few templates.
  • 546: Article without categories. It's currently disabled on enwiki. It requires configuration (see frwiki for an example) to define categorizing templates to avoid false positives.
  • 547: Empty list item. It's currently disabled on enwiki but a dump analysis has been created.
  • 548: Punctuation in link. It detects links with the text ending by a punctuation (which usually should be after the link). It's currently disabled on enwiki but a dump analysis has been created. Configuration could be added (see frwiki for an example) to define links that should be ignored (like [[Comma|,]]).
  • 549: Split link. It detects links that are split in several parts. It's currently disabled on enwiki but a dump analysis has been created.

If you want, some of the errors can be activated here or on other wikis. --NicoV (Talk on frwiki) 11:52, 30 March 2020 (UTC)

Is there a page that has links to all recent dumps? --Guy Macon (talk) 13:32, 30 March 2020 (UTC)
To the dump analysis? You can use the Special page All pages with prefix to list all the analysis pages. There's also Wikipedia:WikiProject Check Wikipedia/WPC all but it's too big so it currently fails to display (visible in edit mode), and doesn't contain the latest additions. --NicoV (Talk on frwiki) 14:10, 30 March 2020 (UTC)


I've added 550: Link without text. It's currently disabled on enwiki but a dump analysis has been created. --NicoV (Talk on frwiki) 17:39, 14 April 2020 (UTC)


I've added 551: Empty line. It's currently disabled on enwiki but a dump analysis should be created soon. --NicoV (Talk on frwiki) 10:17, 21 April 2020 (UTC)

Error 61 at rowiki

Please consider disabling error 61 on Romanian Wikipedia, because as per Romanian guideline on footnotes they are allowed before punctuation, and used as such, including in featured articles (e.g. ro:Jigoku Shoujo, ro:Franz Kafka). Rule on rowiki is just no not mix the styles in one single article. Footnotes after punctuation is still the style used in the majority of articles. Gikü (talk) 16:52, 21 April 2020 (UTC)

@Gikü: I have disabled #61 with this edit. I only changed one line: error_061_prio_rowiki=0. The page save changed a lot of other stuff automatically. --Bamyers99 (talk) 17:23, 21 April 2020 (UTC)

New checking feature?

I wasn't sure where to go, so I've cross-posted this at Wikipedia talk:AutoWikiBrowser.


I made a request at WP:BOTR the other day, and the discussion proceeded as follows:

I think I made this kind of request several years ago, but I can't find it in the archives.

Occasionally people add text like [citation needed] or (reference needed) to articles, and these articles don't end up in maintenance categories because they're plain text instead of templates. Could someone write a bot that would go around making edits like this, or could an existing maintenance-bot operator add this task? I'm guessing that it would be rather simple — give it a list of phrases, tell it to look for them inside parentheses and brackets, and let it loose. Of course, this isn't a one-time problem, so if this is a good idea, it ought to be made an ongoing task. Nyttend backup (talk) 16:35, 20 April 2020 (UTC)

It's probably best to do this with AWB or another semi-auto tool, as there are some legitimate uses of "[citation needed]" in pages (on this page, citation needed, and links to citation needed for example). [7] is a good search term for the first one, then can regex search-and-replace \[?\[citation needed\]\]? with {{subst:cn}}. --AntiCompositeNumber (talk) 15:49, 22 April 2020 (UTC)
@Nyttend backup:   Doing... via AWB. GoingBatty (talk) 12:58, 23 April 2020 (UTC)
@Nyttend backup:   Done - except for those inside comments. GoingBatty (talk) 16:53, 23 April 2020 (UTC)
Another one that could be picked-up is (dead link) or [dead link]. Keith D (talk) 14:29, 23 April 2020 (UTC)
Hi, thanks for the help! I know there are some general maintenance tasks that AWB operators tend to look for. How do I ask that this kind of fix be added to their task list? Nyttend backup (talk) 19:13, 23 April 2020 (UTC)
@Nyttend backup: Seems like a good question for Wikipedia talk:AutoWikiBrowser or Wikipedia talk:WikiProject Check Wikipedia. GoingBatty (talk) 22:38, 23 April 2020 (UTC)
@Keith D:   Doing... but many can be deleted if there's already an archive-url or the link can be fixed. GoingBatty (talk) 01:40, 24 April 2020 (UTC)

Is it possible for your project to start looking for typed-out messages [citation needed] as one of the routine checks that you do? Thank you. Nyttend backup (talk) 18:06, 24 April 2020 (UTC)

Reference in heading?

Is there a scan for reference contained in a heading? RJFJR (talk) 00:55, 18 April 2020 (UTC)

Hello RJFJR. AFAIK, not by Check Wiki. I've added such error a long time ago, as CW Error #504, to WPCleaner. I don't remember what level of testing was done on that error, so it may have a lot of false positives or problems. If you're interested, I can try adding it to the dump analysis performed by WPCleaner (it will be at Wikipedia:CHECKWIKI/WPC 504 dump). --NicoV (Talk on frwiki) 10:11, 21 April 2020 (UTC)
Thank you. I'll take a loom at the dump when it is there. RJFJR (talk) 15:51, 21 April 2020 (UTC)
Hello RJFJR. Wikipedia:CHECKWIKI/WPC 504 dump has been generated. --NicoV (Talk on frwiki) 06:16, 24 April 2020 (UTC)
Works great. Thank you. RJFJR (talk) 18:42, 24 April 2020 (UTC)

Checkwiki request

Hi, Can we include mentioned wikis on Check Wikipedia tools

  • bnwiki
  • bnwikibooks
  • bnwikisource
  • bnwikivoyage
  • bnwiktionary
  • commonswiki
  • mediawikiwiki
  • metawiki
  • wikidatawiki

Warm Regards, ZI Jony (Talk) 13:08, 27 April 2020 (UTC)

@ZI Jony: I think the first step is to create a configuration page in each wiki (like Wikipedia:WikiProject Check Wikipedia/Translation for enwiki, or fr:Projet:Correction syntaxique/Traduction for frwiki). It enables to configure the errors (which one are active, what is their name and description, and other parameters). I advise to start with one wiki to see the result. --NicoV (Talk on frwiki) 14:03, 27 April 2020 (UTC).
And if you're interested in tools to help fixing the issues detected by Check Wikipedia, you can have a look at WPCleaner. --NicoV (Talk on frwiki) 14:07, 27 April 2020 (UTC)
NicoV, we already translated bn:উইকিপিডিয়া:WikiProject Check Wikipedia/Translation, if https://tools.wmflabs.org/checkwiki/ detect issues then we can fix via AWB, WPCleaner or AF. Warm Regards, ZI Jony (Talk) 14:17, 27 April 2020 (UTC)
@ZI Jony: bnwiki has been included, see here. Checkwiki is for checking articles of language specific wiki projects, not commonswiki, mediawikiwiki, metawiki, wikidatawiki. --Bamyers99 (talk) 23:58, 27 April 2020 (UTC)

CheckWiki marks many articles as false positives

Articles are marked as missing <references /> tag (list 3) on cswiki even though they contain the following:

<references>
<ref name="abc">def</ref>
</references>

See for example the following page: cs:Klenový troják v Hřebenech. --Dvorapa (talk) 11:16, 30 April 2020 (UTC)

@Dvorapa: This has now been fixed to allow space(s) after the references tag. ie. <references >. --Bamyers99 (talk) 15:32, 30 April 2020 (UTC)

"Specific Interface" - last run in 2017!?

"A specific interface – here for the English project –" ... it reports the scan being run the last time in 2017 ?!?! CommanderWaterford (talk) 06:58, 26 May 2020 (UTC)

@CommanderWaterford: The semi-monthly Wikipedia dump files for enwiki and dewiki are not scanned due to their large file size which causes a long run time. enwiki, dewiki and some other wikis have their changed articles scanned daily. --Bamyers99 (talk) 21:21, 26 May 2020 (UTC)
Tnx Bamyers99 for clarifying.CommanderWaterford (talk) 21:24, 26 May 2020 (UTC)

New requirements for user signatures

The Editing team is sorting out the upcoming mw:New requirements for user signatures. This should ultimately reduce your workload, if you happen to clean up talk pages. But even if you don't, I expect you all to get more questions, especially along the lines of "How do I make my sig look the same, only without screwing up the page?" You can check the Phab task (especially the dev's recent comment at the end) for more details than are currently on wiki. Whatamidoing (WMF) (talk) 01:28, 30 May 2020 (UTC)

Request for addition of error

Hello!

Please add duplicate headers as an error to fix pages like Archdeacon of Halifax, Tahara Castle, Currant Creek (Juab and Utah counties, Utah) etc. I couldn't find more in the (Article:) namespace but in the File: namespace there are a lot of pages with duplicate headers, see for example file: insource:/== ?Summary ?==.== ?Summary ?==|== ?Licensing ?==.== ?Licensing ?==/ with 3922 entries.Jonteemil (talk) 08:36, 21 April 2020 (UTC)

Hello Jonteemil. There's already CW Error #92, but it's currently disabled for enwiki. If needed, I can add it to WPCleaner's dump analysis (it will be at Wikipedia:CHECKWIKI/WPC 092 dump). Unfortunately, both of them will probably be restricted to main namespace. --NicoV (Talk on frwiki) 08:56, 21 April 2020 (UTC)
@NicoV: I see. Thanks for a fast response. Please add it to the dump analysis if it isn't that troublesome. How come it won't work in the file namespace?Jonteemil (talk) 09:10, 21 April 2020 (UTC)
Btw, there is also yet another error I've found. A very minor one so shouldn't be run by itself. It is all of these that should have the : removed since the template adds it automatically. Now the rendering is two :: after eachother.Jonteemil (talk) 09:14, 21 April 2020 (UTC)
@Jonteemil: It shouldn't be a problem to add it to the dump analysis, but as I've never tried #92 with the dump analysis, it may have false positives... Probably in the next few days, as a new dump should be available soon. The dump analysis is currently restricted to the main namespace (like Check Wiki), I may see later to extend it to the File namespace. --NicoV (Talk on frwiki) 09:49, 21 April 2020 (UTC)
@NicoV: Perfect, thanks. How about the {{ping}} thing? Can/should it be added to the list? Or is the list only for problems in the (Article) namespace?Jonteemil (talk) 10:01, 21 April 2020 (UTC)
@Jonteemil: The ping thing should be discussed in a separate thread, but CW is really designed for the Article namespace, not for talk pages: fixing errors in talk pages isn't appreciated by everyone, so it's usually better to leave the errors (unless, I think, if they pollute categories that are filled automatically by MW like Special:LintErrors/missing-end-tag... but I'm not even sure there's a consensus about that on enwiki). --NicoV (Talk on frwiki) 10:07, 21 April 2020 (UTC)
@NicoV: Okay, thanks a lot!Jonteemil (talk) 10:12, 21 April 2020 (UTC)
@Jonteemil: Wikipedia:CHECKWIKI/WPC 092 dump has been generated. --NicoV (Talk on frwiki) 06:16, 24 April 2020 (UTC)

@NicoV: Thanks! Although though it was only going to list all duplicate headers seperated by \n i.e. a line. All of the ones in the dump aren't incorrect uses. Some are correct I think. Can you exclude everyone that are seperated by at least 3 \n?Jonteemil (talk) 11:06, 24 April 2020 (UTC)

@Jonteemil: For the moment, I can't keep only the ones with empty lines between them, but my bot could fix them automatically (WPCleaner is already able to do that, I just have to request an authorization for my bot). It also fixes cases where consecutive chapters are doubled with their contents too. --NicoV (Talk on frwiki) 12:15, 24 April 2020 (UTC)
@NicoV: That would be brilliant, thank you!Jonteemil (talk) 12:17, 24 April 2020 (UTC)
@Jonteemil: I've requested approval to run my bot on #92. I'm also generating the list again to include also File: and report only consecutive duplicate headers (but there can be text between). --NicoV (Talk on frwiki) 19:57, 25 April 2020 (UTC)
@NicoV: Thanks a lot! Appreciated.Jonteemil (talk) 19:59, 25 April 2020 (UTC)
@Jonteemil: New list has been generated, my bot would be able to modify more than 5.000 pages on the 12.000 pages reported. --NicoV (Talk on frwiki) 09:14, 26 April 2020 (UTC)
@NicoV: Hi again! Why did the bot remove so many pages?Jonteemil (talk) 20:59, 5 May 2020 (UTC)
@Jonteemil: It's because I run the entire dump analysis, and for the moment I didn't include the File: namespace in it. I'm testing it on frwiki first for all pages, before using it on enwiki also. And as I'm still waiting for approval on my request to run my bot, I will add again the File: namespace once I've run the automatic fixing. --NicoV (Talk on frwiki) 10:21, 6 May 2020 (UTC)
@Jonteemil: The task was approved. I've already run my bot on the main namespace, it's going now through the File: namespace. It will require probably a dozen hours to complete. Later, I will re-run a dump analysis for you to see what's left after the bot has done what it could. --NicoV (Talk on frwiki) 14:34, 27 May 2020 (UTC)

@NicoV: Awesome, thanks for that!Jonteemil (talk) 16:53, 27 May 2020 (UTC)

@NicoV: Found an error: Here the last header should be kept, not the first.
@Jonteemil: It's too complex, it keeps the header with the longer content, so content doesn't end up under another section. --NicoV (Talk on frwiki) 19:56, 27 May 2020 (UTC)
@Jonteemil: Wikipedia:CHECKWIKI/WPC 092 dump has been updated. I'm running again the bot for automatic fixing, but only a few articles should be modified now, the rest needing human attention. --NicoV (Talk on frwiki) 08:26, 29 May 2020 (UTC)
@Jonteemil: In fact, my bot seems to find several dozen pages in File: with the same problem (doubled "Summary" title) for images created in the last month. As the error seems to be exactly the same for several editors, maybe there's a problem in the process to upload new images on enwiki? --NicoV (Talk on frwiki) 08:50, 29 May 2020 (UTC)
@NicoV: Are most of the doubled summary headings from the last month? In that case that's very weird. Perhaps you'sre right that it is some upload script that has malfunctioned.Jonteemil (talk) 20:53, 29 May 2020 (UTC)
The 105 edits performed after the last update of the dump analysis are all for problems which happened after April 20th, and they all seem to have the exact same problem (doubled "Summary" title). So, yes, I think there's a bug in the upload process somewhere... --NicoV (Talk on frwiki) 08:19, 30 May 2020 (UTC)

False positive on 03 with {{rfs}}

It's a redirect to reflist, should be added to the whitelist. Jerod Lycett (talk) 01:59, 21 June 2020 (UTC)

@Jerodlycett: rfs has been added with this edit. --Bamyers99 (talk) 15:10, 21 June 2020 (UTC)

False positive with Quantitative genetics and error 98

I have searched through this and I can't seem to find what is triggering it even. Jerod Lycett (talk) 02:42, 21 June 2020 (UTC)

These errors are much easier to find with User:PerfektesChaos/js/lintHint. Ping me if you need help setting it up. – Jonesey95 (talk) 04:19, 21 June 2020 (UTC)
I use WPCleaner. Jerod Lycett (talk) 05:31, 21 June 2020 (UTC)
@Jerodlycett: Jonesey95 fixed the invalid <sub|> tag in the article with this edit. --Bamyers99 (talk) 15:19, 21 June 2020 (UTC)

False positive on CW Error #2

Per the discussion over on WPCleaner, it seems that a <br> with a newline in it is just fine, but it shows up in the list. It's both valid HTML and handled fine by the parser. Jerod Lycett (talk) 19:57, 3 January 2020 (UTC)

@Jerodlycett: I have refined the #2 check to allow any whitespace (newline, tab, etc.) between <br and >. --Bamyers99 (talk) 02:00, 4 January 2020 (UTC)
@Bamyers99: This either got reverted or was not fully fixed as it's showing up at American Experience. Jerod Lycett (talk) 06:14, 22 June 2020 (UTC)
@Jerodlycett: Fixed. This was not a regression. Just more fun with the 8 <br> specific regular expressions. For the code changes see here (GitHub). --Bamyers99 (talk) 18:31, 22 June 2020 (UTC)

HTML assistance

I'm not even sure where to begin with this page to determine what is false and what is true. Jerod Lycett (talk) 02:32, 23 June 2020 (UTC)

Can you be more specific about your concerns? – Jonesey95 (talk) 04:30, 23 June 2020 (UTC)
There are several things showing up, mainly in {{code}}. There are also three uses of h# tags as examples. I think all the ones in the template are false positives, but I'm not sure about the header examples. Jerod Lycett (talk) 07:09, 23 June 2020 (UTC)

Checking for accessibility

Per a request by Guy Macon, I'm reaching out to my fellow editors here to encourage checking for accessibility issues. Four in particular that are on my radar:

Additionally, there are so many uses of <font>...</font> in signatures. :/ ―Justin (koavf)TCM 22:48, 29 June 2020 (UTC)

For table semantics, there would need to be new WP:CWERRORS for those. I am also not sure what the issue is with the scope. For Alt, CW Error #30 exists but is turned off. For color issues, I am not sure if that would be possible to detect programmatically. Same with the misuse of small. We don't generally touch user signatures. (Nor talk pages.) Jerod Lycett (talk) 04:28, 30 June 2020 (UTC)
The problem with scope is that columns should have scope="col" and new rows should have scope="row" but they frequently don't. Semi-automated checking of color contrast is not a trivial thing but something that probably can be done with some false negatives. There are many color contrast tools that can give you an idea of how to think of this (I am not smart enough to solve this problem except for manually). ―Justin (koavf)TCM 21:32, 30 June 2020 (UTC)
Is there a consensus established that tables should have those scopes set? Jerod Lycett (talk) 19:26, 1 July 2020 (UTC)

Domain name change - tools.wmflabs.org/checkwiki => checkwiki.toolforge.org

The tools.wmflabs.org domain is being replaced with toolforge.org. Tools are also being moved from a sub-directory to a subdomain. tools.wmflabs.org/checkwiki is moving to checkwiki.toolforge.org. The new domain is operational and CheckWiki appears to operate correctly on it. Sample new url. The old domain will eventually redirect to the new domain. For more information see wikitech:News/Toolforge.org. --Bamyers99 (talk) 17:20, 23 June 2020 (UTC)

The old domain is now redirecting to the new domain. --Bamyers99 (talk) 14:07, 7 July 2020 (UTC)

Add Moroccan Arabic Wikipedia (arywiki)

Hi, @Bamyers99:. Could you, please, add the Moroccan Arabic Wikipedia (arywiki) to the list of supported projects? I am interested in running my bot to fix errors on this wiki, so, it would help me if it was added to Checkwiki. This is a small wiki, so, it won't take a lot of resources to process its dump. Local translation page is ary:ويكيپيديا:WikiProject Check Wikipedia/Translation. Thank you. --Meno25 (talk) 12:31, 12 August 2020 (UTC)

@Meno25: New report is here. --Bamyers99 (talk) 16:50, 12 August 2020 (UTC)
@Bamyers99: Thank you   --Meno25 (talk) 16:55, 12 August 2020 (UTC)

Heading hierarchy fixes

Hi folks. @WikiCleanerBot: is currently fixing heading hierarchies. Are you changes like [8] and [9] intended? If so, what was wrong about heading hierarchy that has been fixed? Robby.is.on (talk) 15:31, 13 August 2020 (UTC)

@NicoV: I hope this is the right place to inquire. Robby.is.on (talk) 15:41, 13 August 2020 (UTC)
@Robby.is.on:. This is done to apply WP:MOSHEAD: no title levels should be skipped to avoid accessibility issues. --NicoV (Talk on frwiki) 19:37, 13 August 2020 (UTC)
@NicoV: Oh, I hadn't seen that levels had been skipped but I do now. Thanks, Robby.is.on (talk) 07:03, 14 August 2020 (UTC)

AWB not making list

Hi. Even though I am very good at using AWB, I've never worked with checkwiki before. Since last 3-4 days, I have been trying to make list using the source "checkwiki error (number)". But I am getting errors everytime. I tried inputting different keywords in the source/error number bar. When I enter 2, 02, or 50; I get the error "The remote server returned an error: (308) Permanent Redirect." I am using AWB version 6.1.0.1 with default settings. I haven't tried the URL method as I couldn't find the URL(s) to be pasted. Any suggestions on how to create a list using error number or URL method will be appreciated a lot. Regards, —usernamekiran (talk) 13:13, 16 August 2020 (UTC)

found the URLs, but still noting on the checkwiki error (number). —usernamekiran (talk) 13:26, 16 August 2020 (UTC)
@Usernamekiran: Do you know what URL AWB is using? Recently, a domain migration has been done for toolforge servers (see a few posts above), so new URL are in use. Old URL are now redirects (#308), maybe AWB doesn't handle well the redirects? I think you should file a bug report for AWB about this, they either have to fix the handle of #308 redirects or change the URL for CheckWiki they're using. --NicoV (Talk on frwiki) 13:33, 16 August 2020 (UTC)
@NicoV: I found the URLs on https://checkwiki.toolforge.org/ → enwiki → (priority of error solving) → (error description) → list of pages with the particular error. On that list page, there is an option "list for bots". The URL of the "list for bots" can be used to "make list" by AWB under "checkwiki error". I will soon let the AWB guys know about this issue. —usernamekiran (talk) 13:53, 16 August 2020 (UTC)

Last Update 2019-09-08 dewiki

Moin Moin together, in the first screen of the tool you could see, that the last update for dewiki was on 2019-09-08. Could somebody have a look on, if the is any problem and perhaps fix it? Regards --Crazy1880 (talk) 05:55, 12 September 2020 (UTC)

@Crazy1880: The last dump file scan was on 2018-11-01. The dump file is not scanned for dewiki and enwiki because the files are large and take too long to run. Updated articles are scanned daily as indicated by the last update date of 2020-09-08. --Bamyers99 (talk) 14:36, 12 September 2020 (UTC)
Moin Moin Bamyers99, ahh, I see. I'm waiting. Regards --Crazy1880 (talk) 13:39, 13 September 2020 (UTC)
Moin Moin Bamyers99, do you have a feeling how long the update could take? Regards --Crazy1880 (talk) 19:46, 23 September 2020 (UTC)
@Crazy1880: Sorry, I misunderstand your very first question because you used the date 2019-09-08 instead of 2020-09-08. Now I see that you were asking why the date had not changed for 4 days. After further research, I discovered that the program that updates the totals, which includes the last update field, had failed. I have fixed the totals update program. The daily article scans continued to run normally and no article error data was lost. --Bamyers99 (talk) 20:25, 23 September 2020 (UTC)
Moin, thats great, thanks and sorry for misspelling. Regards --Crazy1880 (talk) 08:38, 24 September 2020 (UTC)

False positive for #61

Hi. I think there's a false positive on 1938 San Antonio pecan shellers strike, CW is reporting an error for {{refn|group=note|Also given as Seligmann.{{sfn|Joseph|2018|p=184}}{{sfn|Kirkpatrick|2018}}}}, but the {{sfn}} is inside a {{refn}} and the comma is outside. For information, I've started running my bot on this list a few weeks ago, and I've reduced it by 20,000 errors: I just finished letter C, so the size will go down a lot in the next weeks  . --NicoV (Talk on frwiki) 16:30, 20 October 2020 (UTC)

And probably same problem on 1912 and 1913 United States Senate elections, CW is reporting an error for {{Cite news |date=January 22, 1913 |title=COLT MADE SENATOR |page=4 |work=[[The New York Times]] |url=https://timesmachine.nytimes.com/timesmachine/1913/01/22/100249881.html?pageNumber=4 |ref={{sfnRef|The New York Times, January 22, 1913}}}},. --NicoV (Talk on frwiki) 16:36, 20 October 2020 (UTC)

There's also a false positive on 1931 China floods, CW is reporting an error for {{sfn|Buck|1932|ps={{Page needed|date=May 2020}}, as cited in {{harvnb|National Flood Relief Commission|1933|p=150}}}}: I think it's the comma after {{Page needed}} that's triggering the error. --NicoV (Talk on frwiki) 16:33, 20 October 2020 (UTC)

@NicoV: Nested templates are too complicated for me, so they are not checked anymore. Re: sfnRef - The code was using the extra templates specified on the configuration page as prefixes, so sfn would also check sfnRef. This has been fixed. For information, 2500 pages with the oldest issues are re-scanned every day, so all pages with issues eventually get re-scanned. --Bamyers99 (talk) 19:55, 20 October 2020 (UTC)

Database maintenance

The database used by Checkwiki is undergoing maintenance since Nov 10. It is in read-only mode. This means that the daily scans can not save recent changes. Fixed issues will not get logged as fixed. Stats will not update. The anticipated return to read-write mode is late Nov 11. --Bamyers99 (talk) 15:48, 11 November 2020 (UTC)

Request for new dump style

For some errors, it is highly unlikely that we can fix them in a timely fashion. My suggestion is to add a dump that is not sorted by alphabetical order, but instead sorted by number of views total or number of views per month. This would allow us to target highly active articles first, where the most good can be done, then going through less popular articles that may only see a few views a month. While I have faith in the WCW community, I don't think it will be possible to complete 182,443 reviews in a timely manner, even if new issues were not created in the meantime. I have done this via the pageviews tool as an example Gsquaredxc (talk) 06:41, 7 November 2020 (UTC)

Hi Gsquaredxc. Sorry for the delay in answering, I didn't see it was related to the dump analysis produced by WPCleaner. Interesting, but I'd like to keep the current lists alphabetically sorted, as it's easier to see the difference between 2 dump analysis. Could you use the pageviews tool to create such list and save it somewhere? If so, the procedure could be added in the paragraph at the beginning of each list of errors. --NicoV (Talk on frwiki) 07:09, 23 November 2020 (UTC)

Template:Tooltip, Template:Hover_title, and Template:Abbr

  FYI
 – Pointer to relevant discussion elsewhere.

Please see: Wikipedia:Templates for discussion/Log/2020 December 3#Template:Hover title and Template:Tooltip

Summary:

  • {{Abbr}} (a wrapper for <abbr>...</abbr>) has long been abused for non-abbreviation markup (against the HTML specs).
  • We had a template, {{Tooltip}}, with <span>...</span> for non-abbreviation use, but it was "merged" (not really) and redirected to {{Abbr}}.
  • The redir was then deprecated (for the reason mentioned above), but the community ignored the deprecation.
  • In the interim, {{Hover title}} was created to do the same thing, but with backwards parameters (and some additional features).
  • Both the {{Tooltip}} then-redirect and {{Hover title}} template have been transcluded in tens of thousands of articles, mainly via infoboxes and other templates.
  • I created a new {{Tooltip}} template, with all the features of {{Hover title}} but preserving the {{Abbr}} parameter order (to not break deployed translcusions).
  • The TfM linked above would merge away {{Hover title}}, but it's going to require flipping the |1= and |2= parameters of its extant instances.
  • Oh, and the documentation would need updating after merger, of course.

 — SMcCandlish ¢ 😼  00:26, 4 December 2020 (UTC)

More database downtime : 1700 UTC on 16 December 2020

The CheckWiki database is going down for more maintenance on 16 December 2020 at 1700 UTC for a couple of hours. Same effects as last time. Announcement details. --Bamyers99 (talk) 19:56, 9 December 2020 (UTC)

Bug in detection of wrong ISBNs?

Hi, the german list of ISBN-13 with wrong Checksum (ID 73) [10] detects more and more false positives. It claims a certain ISBN-13 as having the wrong checksum, but the website of the publisher (famous like "Springer") declares these ISBNs as correct, same for the German National Library (Deutsche Nationalbibliothek, DNB). I'm not sure whether this happens only for E-Book ISBNs. Examples are

Is there a bug in the algorithm? Any other reason? Many thanks --Bicycle Tourer (talk) 04:56, 9 December 2020 (UTC)

@Bicycle Tourer: Evidently the ISBN agency keeps updating the list of valid ISBN Ranges. I have updated CheckWiki to use the current range list (previous list was from 2014). CheckWiki is now reporting 1 instead of 4 invalid ISBNs for de:Helmut Günther (Physiker) --Bamyers99 (talk) 19:41, 9 December 2020 (UTC)
@Bamyers99: Many thanks for updating the valid ISBN ranges. For the two examples given above I corrected the articles for the remaining finding, we have to wait for the next run in the forthcoming night. However, the number of findings for ISBN-13 wrong checksum (ID 73) jumped from about 35 to now 149 last night. I tried to fix the first five entries, with the following results:
  • de:7. Sinfonie (Beethoven), ISBN 978-8884567895: DNB knows the book ([11] from 1975), but not the ISBN --> Inconclusive, further research needed by myself,
  • de:Abessinienkrieg, ISBN 978-88-943205-2-7: This book exists with exactly this ISBN (DNB: [12]) --> false positive
  • de:Aby Warburg --> Inconclusive
  • de:Agust D (Mixtape), ISBN 9788925565828: Book exists (Library of Congress Catalogue (LCC): [13]) --> false positive
  • de:Alcalá de Henares, ISBN 978-84-16599-76-9: A PDF [14] including a barcode with this ISBN on the backcover, i.e. this ISBN exists and is correct --> false positive
I stop after 3 out of 5 being false positives and jump into a detailed analysis of one of them: 978-88-943205-2-7 (out of de:Abessinienkrieg) shows (based on the ISBN range document given by link above [15]):
  • 978-88: Code for Italy, --> correct
  • 943205: Part of "940000-947999", a shown range inside Italy, --> correct
  • 2: Booknumber (publication element)
  • 7: Checksum: 10 - ((1*(9+8+8+4+2+5) + 3*(7+8+9+3+0+2)) mod 10) = 10 - ((1*36 + 3*29) mod 10) = 10 - ((36 + 87) mod 10) = 10 - (123 mod 10) = 10 - 3 = 7 --> correct
Any ideas? Many thanks and BR --Bicycle Tourer (talk) 16:55, 10 December 2020 (UTC)
Hi. I checked with WPCleaner and none of your 3 examples are detected as incorrect with it. It's also using the list of valid ISBN ranges from the same source, so it doesn't seem to be the reason. --NicoV (Talk on frwiki) 17:53, 10 December 2020 (UTC)
@Bicycle Tourer: I have just now upgraded the 3rd party software (Business::ISBN) that does the ISBN validation. That seems to have gotten rid of the false positives. I have marked the false positives as done in the error list. --Bamyers99 (talk) 19:43, 10 December 2020 (UTC)
Many thanks to all. I will wait for the forthcoming night and the new list. BR Bicycle Tourer (talk) 21:59, 10 December 2020 (UTC)
  Done The lists look good now, so this issue is resolved. Many thanks to all. --Bicycle Tourer (talk) 14:33, 12 December 2020 (UTC)

WCW or User crap?

On 2 Novenber 2020 this edit was made with comment "(v2.03 - WP:WCW project (Category duplication - Template programming element - Spelling and typography))". Are any of the tools used in WCW responsible for the crap created or is it the responsibility of the editor Gsquaredxc? --Sb008 (talk) 16:08, 16 December 2020 (UTC)

See a report of the same problem, above. This is a bug. WPCleaner should not do this. – Jonesey95 (talk) 16:49, 16 December 2020 (UTC)
That leaves the question who's going to correct the page, back to its correct format? From what I read above, WPCleaner only facilitates, however a human editor makes the descision. So is the creator of WPCleaner or the editor going to correct the page? And is the editor in any way informed to think thrice before using this WPCleaner option? --Sb008 (talk) 17:27, 16 December 2020 (UTC)
??? The page was already fixed, half an hour before the above response. – Jonesey95 (talk) 18:44, 16 December 2020 (UTC)
Sb008, this is my mistake, however most likely Lua programming elements are incorrectly used and should be fixed in some way. Example being: {{#expr: 9100 + 3605 + 1300 + 1900 + 4900 + 2864 + 1450 + 1800 + 1300 + 2230 + 1871 + 1400 + 1580 + 1910 + 1450 + 1850 + 1600 + 1380 + 5570 + 6628 + 4476 + 4690 + 1680 + 2100 + 6765 + 5060 + 1750 + 1950 + 1600 + 1985 + 2280 + 1540 + 1810 + 6375 + 1970 + 1000 + 1590 + 2000 + 1700 + 1950 + 2020 + 1100 + 5520 + 6729 + 2460 + 6220 + 1870 + 2442 + 4680 + 1490 + 2000 + 1700 + 8310 + 2680 + 2950 + 2100 + 2500 + 3380 + 1600 + 3826 + 800 + 3430 + 3726 + 1000 + 5870 + 3450 + 1100 + 7356 + 4020 + 600 + 2132 + 1890 + 6920 + 1700 + 6464 + 5090 + 1900 + 2920 + 2200 + 600 + 2120 + 6338 + 3428 + 1980 + 6281 + 5070 + 3962 + 3080 + 1926 + 3520 + 5690 + 6860 + 4240 + 4261 + 8664 + 9624}} Perhaps there is something that can replace the Lua functions that are in direct wiki code, as most people will have issues editing if they have to search code in order to understand how to edit a table. Gsquaredxc (talk) 00:59, 17 December 2020 (UTC)
Gsquaredxc {{#expr: }} is not a Lua programming element but part of the parser Functions. If you think that these functions should be replaced by direct wiki code, we probably have to get rid of the great majority of templates. Furthermore, you say "most people will have issues editing if they have to search code in order to understand how to edit a table." Besides the example you gave above, on the page you changed all tables, defined by easy to read parameter definitions for modules (which actually are written in LUA and can be recognized cause they start with {{#invoke:), into for many people hard to read HTML code. So, you did exactly the opposite of what you claim what should be done. You went from easy to read and edit, to much harder to read and edit. But I guess, just like with templates, we should get rid of all modules (since they're written in LUA) as well. I think it's a much better idea, if you first made sure you understand the effects and consequences of functionalities of edit tools before you actually start using them. You should have stopped after typing "this is my mistake" and not have mentioned bogus arguments in the "however" part. Last, on 28 november you were already notified about your mistake and you did nothing to correct it. --Sb008 (talk) 04:33, 17 December 2020 (UTC)
Sb008, what I'm saying is that it is difficult for people to understand what 9100 + 3605 + 1300 + 1900 + 4900 + 2864 + 1450 + 1800 + 1300 + 2230 + 1871 + 1400 + 1580 + 1910 + 1450 + 1850 + 1600 + 1380 + 5570 + 6628 + 4476 + 4690 + 1680 + 2100 + 6765 + 5060 + 1750 + 1950 + 1600 + 1985 + 2280 + 1540 + 1810 + 6375 + 1970 + 1000 + 1590 + 2000 + 1700 + 1950 + 2020 + 1100 + 5520 + 6729 + 2460 + 6220 + 1870 + 2442 + 4680 + 1490 + 2000 + 1700 + 8310 + 2680 + 2950 + 2100 + 2500 + 3380 + 1600 + 3826 + 800 + 3430 + 3726 + 1000 + 5870 + 3450 + 1100 + 7356 + 4020 + 600 + 2132 + 1890 + 6920 + 1700 + 6464 + 5090 + 1900 + 2920 + 2200 + 600 + 2120 + 6338 + 3428 + 1980 + 6281 + 5070 + 3962 + 3080 + 1926 + 3520 + 5690 + 6860 + 4240 + 4261 + 8664 + 9624 means or why it is there. As for the previous mention, I have been fairly busy since then and didn't believe there were many edits that had this issue. In terms of the common way to do things, people usually use templates instead of invoking directly to Lua. Directly invoking Lua is considered an error on the WCW project. Don't forget to assume good faith, because at least how I interpreted your message was not in that fashion. Gsquaredxc (talk) 06:28, 17 December 2020 (UTC)
Gsquaredxc
  • Good faith: I'm more than happy to assume good faith when people make mistakes. However, when someone is notified about a mistake and does nothing to correct the mistake, it becomes hard to find any good faith in there, especially if he/she presents lame excuses instead. You didn't believe there were many edits that had this issue? So, you only correct your mistakes only if there're a lot of them, and if it aren't that many, somebody else can clean up the mess you caused? Feel free to explain how I should interpret that as good faith. The defailt is to receive an email when somebody leaves a message on your talk page. Maybe you switched off that setting or you use an email which you don't check that often or you simply don't act on those emails. Either way, did nothing because of a choices you made. Again, feel free to explain how I should interpret that as good faith.
  • LUA: once again "{{#expr: }}" is no LUA. So, LUA is not invoked directly. The only thing that's invoked are LUA modules, which is very common.
  • Understanding: Are you telling me people don't understand a simple sommation, but do understand all the HTML code you converted the tables into? I doubt that people who don't understand a sommation, will understand most of the templates and modules. Are you advocating to get rid of everything people might not understand? We should get rid of Wiki code as well. After all, a new editor without foreknowledge of Wiki code, won't understand that either. Should be nice, a plain text Wiki. We can expect of any editor to put some effort in trying to understand whatever code is used on a page. There are plenty of help-pages and if those don't provide an answer, you can always ask someone. If you don't want to make an effort or ask someone, you should limit your edits to the plain text parts.
  • considered an error: Where on the list of errors is anything mentioned about LUA? What's mentioned though is the low priority error "HTML table element". After your action the page was crowded with HTML table elements and that's why the {{cleanup HTML|tags=table, tr, td}} template was placed on the page.
All in all, you made error PERIOD. Something which can and has happened to all of us. However not all who make a mistake, use lame arguments and excuses in an attempt to straighten out what is crooked and to justify their mistakes. It was your choice not to receive/read/act on a message about the mistake on your talk page. Be a man and take responsibility for your mistakes. Not doing so, and you understood that correct, I indeed can't consider as good faith. Put an effort in understanding (result and impact) of the tools you use, or don't use them. I will not respond to any furher discussion because if it isn't clear to you now, it will never be and a discussions like this don't belong on this page. --Sb008 (talk) 17:54, 17 December 2020 (UTC)
Sb008, please try to AGF and be polite, it seems as though you have not assumed any good faith or attempted to be polite. In the past you have not been polite to other editors. Please try to be polite to others on Wikipedia. BangJan1999, Zackmann and HawkAussie have previously asked you to do this in the past.
The error is error 34, which is high priority.
My point is that you seem to be arguing for keeping everything identical, whereas I would like to improve the pages to remove anything that would be confusing or difficult to edit. Yes, my accidental edits that caused issue didn't do that. However, I can guarantee you spent much more time arguing with me about how bad my edits were than it would have taken to revert the THREE edits that were mistakenly created. I went through my edits, out of my over 8500 edits, only 5 edited template programming elements, and out of those 5, only 3 had the issue. You can probably understand why one singular incorrect edit that was told to me on my talk page doesn't mean I need to search through my entire edit history to check if any edit was the same as the singular edit that took 26 days to catch. I was correct, and there were only 3 edits, which is 0.035% of my edits. I have been extraordinarily busy, and didn't really have any time to find and correct the other 2 edits that I didn't even know existed at the time. I don't believe that saying, "Perhaps there is something that can replace the Lua functions that are in direct wiki code" isn't "being a man" as you say. I would kindly request that you would consider if what you say could be interpreted as in good faith. I admitted that my edits were in error but also explained that the pages could be improved. Is this "not good faith"? Gsquaredxc (talk) 22:30, 17 December 2020 (UTC)

Request to change the layout of this project pages and talk page

The current layout makes it very difficult to read in mobile devices. The border fits the screen but text is not getting wrapped up to fit the screen. I have to keep scrolling back and forth horizontally to read a full line. Can anyone please modify this so that is friendly for mobile view? Thanks AVSmalnad77 chat 17:13, 1 January 2021 (UTC)

Greater after br

This is detected as an error:

<br />>

But it is valid. I have tables with br-separated lines and numbers, and there can be a < or > to indicate the number range, e. g. "> 60". This will result in the code :<br />>60 . Suggest to allow this in syntax check. --PM3 (talk) 10:16, 20 February 2021 (UTC)

Hi PM3. Allowing that unusual combination might resolve the false positive that you're experiencing, but it might also allow actual syntax errors to go undetected. When encountering a special circumstance such as this, there are often templates that can avoid the false positive. Do you think Template:Break {{break}} will help in your tables? —Scottyoak2 (talk) 12:30, 20 February 2021 (UTC)
Sure there are workarounds, but you have to think of using them. Probably I will forget that often. ;)
This actually never is a code syntax error but a content semantic error: It will insert a > symbol in the output of the page. Just if you typed any other unwanted character, e.g. a dot or a ?, which could as well accidentally happen after typing a > (an will not be caught by this syntax checker, I assume ...?). But if >> is a common mistyping and there is a significant number of true positives by this check, I would agree that it is helpful. --PM3 (talk) 13:50, 20 February 2021 (UTC)
@PM3: Can you indicate which checkwiki error code this is for and give an example article that is reporting the error. --Bamyers99 (talk) 21:58, 20 February 2021 (UTC)
It recently happened in this dewiki article version. I don't know about the check details. Girus may be able to explain more, he processed the check report. --PM3 (talk) 22:18, 20 February 2021 (UTC)
@PM3: Fixed. --Bamyers99 (talk) 01:03, 21 February 2021 (UTC)
Thanks! --PM3 (talk) 09:45, 21 February 2021 (UTC)