Wikipedia:Bot requests

From Wikipedia, the free encyclopedia
  (Redirected from Wikipedia:BOTREQ)
Jump to: navigation, search

This is a page for requesting work to be done by bots per the bot policy. This is an appropriate place to simply put ideas for bots. If you need a piece of software written for a specific article you may get a faster response time at the computer help desk. You might also check {{Botcats}} to see if the bot you are looking for already exists, in which case you can contact the operator directly on his or her talkpage.

If you have a question about one particular bot, it should be directed to the bot owner's talk page or to the Bot Owners' Noticeboard. If a bot is acting improperly, a note about that should be posted to the owner's talk page and to the Administrators' Noticeboard. A link to such a posting may be posted at the Bot Owners' Noticeboard.

If you are a bot operator and you complete a request, note what you did, and archive it. {{BOTREQ}} can be used to give common responses, and to make it easier to see at-a-glance what the response is.

There are a number of common requests which are regularly denied, either because they are too complicated to program, or do not have consensus from the Wikipedia community. Please see Wikipedia:Bots/Frequently denied bots for a list of such requests, and ensure that your idea is not among them.

If you are requesting that a bot be used to add a WikiProject banner to the talkpages of all articles in a particular category or its subcategories, please be very careful to check the category tree for any unwanted subcategories. It is best to give a complete list of categories that should be worked through individually, rather than one category to be analyzed recursively. Compare the difference between a recursive list and a properly vetted one.

Please add your bot requests to the bottom of this page.
Make a new request


Removal of duplicated citations[edit]

I suggest a bot that can remove duplicated citations. If you look at the source code, you can see what I mean by "duplicated citations". Qwertyxp2000 (talk) 23:41, 6 April 2015 (UTC)

Markup Renders as
====Without duplicated citations===
Lorem ipsum dolor sit amet, consectetuer adipiscing elit.<ref name="random thingy" group="example ref1">[ Random citation] Google. Retrieved at "random date".</ref> Aenean commodo ligula eget dolor. Aenean massa. Cum sociis natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Donec quam felis, ultricies nec, pellentesque eu, pretium quis, sem. Nulla consequat massa quis enim. Donec pede justo, fringilla vel, aliquet nec, vulputate eget, arcu. In enim justo, rhoncus ut, imperdiet a, venenatis vitae, justo. Nullam dictum felis eu pede mollis pretium. Integer tincidunt. Cras dapibus..<ref name="random thingy" group="example ref1" /> Vivamus elementum semper nisi. Aenean vulputate eleifend tellus. Aenean leo ligula, porttitor eu, consequat vitae, eleifend ac, enim. Aliquam lorem ante, dapibus in, viverra quis, feugiat a.

====Dummy refs====
{{reflist|group="example ref1"}}

{{tick}} This is acceptable

===With duplicated citations===

Lorem ipsum dolor sit amet, consectetuer adipiscing elit.<ref group="example ref2">[ Random citation] Google. Retrieved at "random date".</ref> Aenean commodo ligula eget dolor. Aenean massa. Cum sociis natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Donec quam felis, ultricies nec, pellentesque eu, pretium quis, sem. Nulla consequat massa quis enim. Donec pede justo, fringilla vel, aliquet nec, vulputate eget, arcu. In enim justo, rhoncus ut, imperdiet a, venenatis vitae, justo. Nullam dictum felis eu pede mollis pretium. Integer tincidunt. Cras dapibus..<ref group="example ref2">[ Random citation] Google. Retrieved at "random date".</ref> Vivamus elementum semper nisi. Aenean vulputate eleifend tellus. Aenean leo ligula, porttitor eu, consequat vitae, eleifend ac, enim. Aliquam lorem ante, dapibus in, viverra quis, feugiat a.

====Dummy refs====
{{reflist|group="example ref2"}}

{{cross}} This is not acceptable 

Without duplicated citations

Lorem ipsum dolor sit amet, consectetuer adipiscing elit.[example ref1 1] Aenean commodo ligula eget dolor. Aenean massa. Cum sociis natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Donec quam felis, ultricies nec, pellentesque eu, pretium quis, sem. Nulla consequat massa quis enim. Donec pede justo, fringilla vel, aliquet nec, vulputate eget, arcu. In enim justo, rhoncus ut, imperdiet a, venenatis vitae, justo. Nullam dictum felis eu pede mollis pretium. Integer tincidunt. Cras dapibus..[example ref1 1] Vivamus elementum semper nisi. Aenean vulputate eleifend tellus. Aenean leo ligula, porttitor eu, consequat vitae, eleifend ac, enim. Aliquam lorem ante, dapibus in, viverra quis, feugiat a.

Dummy refs

  1. ^ a b Random citation Google. Retrieved at "random date".

YesY This is acceptable

With duplicated citations

Lorem ipsum dolor sit amet, consectetuer adipiscing elit.[example ref2 1] Aenean commodo ligula eget dolor. Aenean massa. Cum sociis natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Donec quam felis, ultricies nec, pellentesque eu, pretium quis, sem. Nulla consequat massa quis enim. Donec pede justo, fringilla vel, aliquet nec, vulputate eget, arcu. In enim justo, rhoncus ut, imperdiet a, venenatis vitae, justo. Nullam dictum felis eu pede mollis pretium. Integer tincidunt. Cras dapibus..[example ref2 2] Vivamus elementum semper nisi. Aenean vulputate eleifend tellus. Aenean leo ligula, porttitor eu, consequat vitae, eleifend ac, enim. Aliquam lorem ante, dapibus in, viverra quis, feugiat a.

Dummy refs

  1. ^ Random citation Google. Retrieved at "random date".
  2. ^ Random citation Google. Retrieved at "random date".
N This is not acceptable
@Qwertyxp2000: AWB's general fixes will do this - see the page for more details. GoingBatty (talk) 01:10, 7 April 2015 (UTC)
GoingBatty, thank you for finding the right page. I will soon be changing the {{Duplicated citations}} tag. Qwertyxp2000 (talk) 01:15, 7 April 2015 (UTC)
@Qwertyxp2000: You might want to have the template link to WP:REFNAME instead of the AWB page. GoingBatty (talk) 01:20, 7 April 2015 (UTC)
@Qwertyxp2000: You might want to have a comment in the documentation saying that AWB may be used to fix the issue, and provide the link to the AWB page. GoingBatty (talk) 01:23, 7 April 2015 (UTC)
Why cannot you do this all? Then I can see what you are thinking. Qwertyxp2000 (talk) 01:31, 7 April 2015 (UTC)
@Qwertyxp2000: Apparently some people think that duplicate citations are acceptable. GoingBatty (talk) 01:41, 7 April 2015 (UTC)
Looking at the second scenario, if I have referenced the first and last sentences of a paragraph to the same source but not the middle of it, or per haps the middle is cited to another source, then if someone comes along and removes a "duplicate" cite, I would revert that as vandalism. We encourage people to use inline citation and multiple sources, but we don't limit people to only citing one statement from each source they use. ϢereSpielChequers 05:58, 7 April 2015 (UTC)
I think you are misunderstanding, WereSpielChequers. No one is saying a statement can only be referenced once. Qwertyxp2000 is wanting a bot to fix references which are duplicated (rather than referenced twice or more). Duplicated references produce two entries to the same thing in the list of references, whereas a reference used multiple times will have one entry with multiple uses (the little "^ a b" you see next to the example reference in the first example). I recently manually combined a bunch of duplicate references here (I also normalized the references so they could be referred to multiple times). Maybe that will help clarify things. ···日本穣? · 投稿 · Talk to Nihonjoe · Join WP Japan! 18:07, 24 May 2015 (UTC)
So is this a good idea for a bot or is it already fulfilled by AWB? Because I am looking to start working on a bot (something I've been putting off for two years). :P Sn1per (talk)(edits) 13:51, 27 June 2015 (UTC)
@Sn1per: AWB's general fixes will also do this in some circumstances - see the page for more details. GoingBatty (talk) 22:57, 28 June 2015 (UTC)
@GoingBatty: I see this problem a lot though and perhaps it would be a good idea to have it actively fixed by a bot to take some work off of AWB users? Sn1per (talk)(edits) 15:07, 29 June 2015 (UTC)
@Sn1per: I don't object to a bot task to do this using AWB's rules. How would you create a list of articles to be fixed? GoingBatty (talk) 15:38, 29 June 2015 (UTC)
@Sn1per and GoingBatty: A database scan for <ref>([^\<]+)</ref>.+<ref>\1</ref> will find about 20,000 candidates. The first 1% or so are listed at User:John of Reading/Sandbox. But remember that the AWB general fixes will only combine duplicate citations if the article already has at least one named reference, to avoid changing the citation style (AWB documentation). -- John of Reading (talk) 16:09, 29 June 2015 (UTC)
@John of Reading: Thanks John! You might want to tweak your regex to also include named references. GoingBatty (talk) 16:28, 29 June 2015 (UTC)
@GoingBatty: That would take more than a "tweak", but I'll think about it. If two references have the same name, the software will use only the first definition whether or not the definitions are identical. So the search would have to be for references that have identical content but different names, or one named and one unnamed. -- John of Reading (talk) 16:40, 29 June 2015 (UTC)
@John of Reading: Thanks for the regex, I tend to be terrible at those :P I would assume that my bot should follow the same behavior as AWB to comply with the same policy? Sn1per (t)(c) 22:54, 29 June 2015 (UTC)
@Sn1per: Definitely, and even then you may run into objections. See this 2011 thread. Hint: it's surprisingly long. -- John of Reading (talk) 06:04, 30 June 2015 (UTC)
@John of Reading: I think I tweaked the regex, not sure if it works all the time but should be accurate enough to quickly select pages for further scrutiny. Are there any obvious errors? <ref(.|\n)*?>([^\<]+)<\/ref>.+<ref(.|\n)*?>\2<\/ref> Sn1per (t)(c) 18:18, 2 July 2015 (UTC)
(still working on improving the regex) Sn1per (t)(c) 18:33, 2 July 2015 (UTC)
@Sn1per: The ".+" in the middle will only work if the regex is run with the "singleline" option turned on - the dot/period needs to match newlines - so the "(.|\n)" can be simplified to just ".". That's <ref.*?>([^\<]+)<\/ref>.+<ref.*?>\1<\/ref>. -- John of Reading (talk) 18:47, 2 July 2015 (UTC)


@John of Reading: Thanks for the advice. Here is an improved regex: <ref([^\>]*)?>([^\<]*)</ref>.*?<ref(?!\1)[^\>]*?>\2</ref> It (should) be able to find two refs, where at least one has no attributes i.e. name="", or if both have different attributes. Sn1per (t)(c) 19:15, 2 July 2015 (UTC)
(note that I am using python regexs, where a "." doesn't seem to match newlines) Sn1per (t)(c) 19:18, 2 July 2015 (UTC)
Nevermind, I'm an idiot. Just say your note about singleline mode. Sn1per (t)(c) 19:19, 2 July 2015 (UTC)
@Sn1per: Yes, that regex does the job - neat! Now, a quick reality check: on my laptop it will take 20 hours to run this against a database dump. My latest dump is from mid-May; I think it's not worth tying up the machine for so long to produce a list that is several weeks out of date. But I can produce a partial list of a hundred or so articles for testing purposes fairly easily. -- John of Reading (talk) 19:45, 2 July 2015 (UTC)
Document-properties.svg Coding... Well, it seems like a good idea to me, so I'll start working on it. But if you guys strongly disagree, leave a message on my talk page so I don't waste too much effort. The bot will probably take a few days to work on given that I am new to the field. Sn1per (talk)(edits) 17:16, 28 June 2015 (UTC)
@Sn1per:, Why code it? Why not use the AWB library that already contains the code. - X201 (talk) 15:33, 29 June 2015 (UTC)
@X201: I was thinking of making a pywikibot-based robot based on the WMF Tool Labs servers so that the bot can run all the time, rather than relying on me to run it off my PC, given that the problem is pretty large. Sn1per (t)(c) 22:54, 29 June 2015 (UTC)

Woken up[edit]

  • I poked Sniper about this and he is still working on it. He hopes to have it ready soon.—cyberpowerChat:Online 20:17, 7 September 2015 (UTC)

Redirects to lists, from the things they are lists of[edit]

Please could someone do this:

  1. For every article titled "List of foo"
  2. if the article called "Foo" exists; do nothing
  3. otherwise, create "Foo" as a redirect to "List of foo"

For example, I just created Birds of Tunisia as a redirect to List of birds of Tunisia.

This might usefully be added to a list of monthly cleanup tasks, for new "List of..." articles. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:06, 6 May 2015 (UTC)

Pictogram voting wait.svg Doing... - Though I have messaged WikiProject Lists to check consensus first. Jamesmcmahon0 (talk) 12:54, 7 May 2015 (UTC)
Thank you. Please see also #Century-item redirects, below. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:36, 9 May 2015 (UTC)
Symbol wait.svg BRFA filed - Wikipedia:Bots/Requests for approval#MoohanBOT 8 It is just for this task as I had already generated the list of pages needed and there seems to be no opposition to it. I will have a look at #Century-item redirects in a few days but feel free to jump ahead GoingBatty as that one may be outside of my regex expertise... Jamesmcmahon0 (talk) 11:05, 10 May 2015 (UTC)

It appears that User:Jamesmcmahon0 has dropped this. Can anyone else help, please? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 22:29, 5 July 2015 (UTC)

I see User:Jamesmcmahon0 has been editing again... Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:30, 3 August 2015 (UTC)
  • The BRFA expired, this task is now open for grabs again.—cyberpowerChat:Online 20:20, 7 September 2015 (UTC)

GA Cup[edit]

On the behalf of myself and Figureskatingfan, we are looking at the possibility of having a bot help assist us in the GA Cup. We held the first competition at the end of 2014/beginning of 2015 and after the success of it, we are currently planning a second competition, hoping for it to be a even bigger success. In the first competition, some of the participants expressed their frustration in the how the submission process for their Good article reviews was not very efficient. For the upcoming competition, we were wondering if it would be possible to have a bot scan the Good article nomination page for reviews being conducted by participants and add the appropriate review links to a page.

More specifically, ideally, the bot would scan the nomination page and say BenLinus1214 was reviewing an article, it would add it under the appropriate header.

If anyone is interested in helping us I would be glad to have you on board and will be more than happy to answer any questions!--Dom497 (talk) 23:25, 11 May 2015 (UTC)

@Dom497: Doesn't seem that hard or complicated. I'll test around a bit but I'm not going to promise anything yet. -24Talk 21:39, 28 May 2015 (UTC)
@Negative24: Thank you so much for trying! I would have done it myself but I don't know enough code to do it!--Dom497 (talk) 22:35, 28 May 2015 (UTC)
@Negative24: Hey, just curious to see if you've had any positive results. Thanks again! :) --Dom497 (talk) 02:30, 14 June 2015 (UTC)
@Dom497: Sorry for the delay. I've been exploring everything related to bots and this being my first time even trying to make a bot, things are going a bit slow. I'm going to be going on a month long Wikibreak (related to a project in life that needs my time) and so I may not be able to code anything up before the 2015 GA cup. I'm going to leave this open for anyone to pickup (feel free to do so if you're interested) but I'm not able to fully start on something at the moment. I will resume this project when I have the time and if someone hasn't picked it up by then. Sorry, -24Talk 03:08, 19 June 2015 (UTC)
@Negative24: No problem at all! Thanks for trying! :) --Dom497 (talk) 19:26, 19 June 2015 (UTC)
Symbol question.svg Question: While I am good with botwork, my experience with the GA process is non-existent. It would help to have some clarification. Maybe you can show me a scenario, and show me the diffs that need to be done?—cyberpowerChat:Online 22:02, 26 August 2015 (UTC)
@Cyberpower678: If a GA nomination is being reviewed, the list of noms at WP:GAN lists their name (see this diff). With the GA Cup there are submission pages (see here for mine). On these I put all the articles I'm reviewing using a special template. This currently has to be done manually, and people forget. What I think Dom497 is looking for is a bot that will periodically look at WP:GAN, see if a GA Cup competitor is reviewing an article, and add that article to their respective submission page using the template. Hope that helps, if something isn't clear, just ping me as I'm not watching this page. Wugapodes (talk) 03:46, 28 August 2015 (UTC)
I asked Legoktm if he could do it since he already runs bots in that area, but he says he's too busy, so I guess I'll do it. Keep an eye on this request.—cyberpowerChat:Online 23:02, 28 August 2015 (UTC)
@Dom497 and Cyberpower678: This would definitely be helpful in future cups, but this probably shouldn't be given an extremely high priority. Let's say Cyberpower works like lightning and gets this bot ready in a week from now. At the current rate bot approvals is moving, the approval will likely take another 2-3 weeks. That would give you maybe 3 weeks of use of this bot in the final round, and that's a best case scenario if Cyberpower creates the bot rapidly. It might be best to code and trial this bot during the "offseason" to prepare it for full use in the next cup. That would also be less likely to cause frustration for the contestants due to a late switch in how things are done. This could be trialled by popping any few regular GA reviewers in as "contestants" for the bot and letting it do its thing so it's ready to go out of the gate for the next cup. ~ RobTalk 14:07, 9 September 2015 (UTC)

Accidental template protection[edit]

Occasionally, I've noticed that an article has been mistakenly template-protected. Perhaps a bot could monitor the protection log, and if a page in a namespace other then Template, Module, User, or Wikipedia is template-protected, deliver a "did you mean to do this" message to the protecting admin, i.e.

Hello administrator name. On date you template-protected [[page]] ([ (url-encoded) log]). As template protection is only meant to be used for templates, or other highly transcluded pages, did you perhaps mean to select a different level? Thanks, bot signature

or similar. - Evad37 [talk] 03:54, 17 June 2015 (UTC)

Unsure if necessary - couldn't you just add something to said template? (Just lurking WP:BOTREQ to see what sort of things people want bots to do). E. Lee (talk) 04:59, 17 June 2015 (UTC)
@Elee: Which template are you talking about? And how does adding something to a template fix a wrongly applied protection level? - Evad37 [talk] 05:50, 17 June 2015 (UTC)
For an example of the problem and proposed solution (letting admins know that they may have made a mistake so they can fix it), see User talk:Ponyo#Maithali protection level, or User talk:Black Kite#Farshad Fotouhi‎ protection level - Evad37 [talk] 05:54, 17 June 2015 (UTC)
There are padlock templates added to protected pages. These "sense" if they are incompatible with the protection actually used, I believe, and put the page in a category to be fixed.
Arguably there is something that could be done along these lines.
A list of template protected articles can be found here (currently empty). A bot could check this, and act upon it. All the best: Rich Farmbrough, 21:43, 27 July 2015 (UTC).
Such a bot should also check the queue for move protection. ~ RobTalk 14:11, 9 September 2015 (UTC)

Articles with {{Infobox Journal}} seek bot to ensure redirects are in place[edit]

As discussed at Wikipedia_talk:WikiProject_Academic_Journals#Bot_task?, there are several fairly standard redirects needed to each article in this project using that infobox. The box has parameters for the journal title and it's ISO abbreviation. Citations routinely vary the capitalization, abbreviations, and punctuation of these abbreviations, creating a need for redirects from each common variation to the actual article title (usually the same as the journal title, in sentence case). Is there a bot that might be suited to the task? LeadSongDog come howl! 01:23, 18 June 2015 (UTC)

  • I've obtained the ISO 4 vocabulary to convert, e.g., "European Physical Journal" to "Eur. Phys. J."; it's a spreadsheet-format version of the PDF available at Could you bot-wizards please tell us if such a conversion would be simply too complicated? Thanks! Fgnievinski (talk) 02:30, 30 July 2015 (UTC)
  • Maybe an easier and useful thing to do instead would be to start from Infobox journal's title field (e.g., "European Physical Journal") and its manually-entered abbreviation field ("Eur. Phys. J."), and create the desired redirects: e.g., "European physical journal", "Eur. Phys. J.", "Eur Phys J", "eur phys j", "E. P. J.", "E.P.J.", "E P J", "EPJ". Fgnievinski (talk) 02:45, 30 July 2015 (UTC)

Replacement of Template:Infobox Country World Championships in Athletics[edit]

Hello. Could I hire a bot to substitute all transclusions of {{Infobox Country World Championships in Athletics}}, per the outcome of this TfD? Alakzi (talk) 13:12, 20 June 2015 (UTC)

Same with {{Infobox China station}} and {{Infobox Japan station}}, but using the sandbox version. Alakzi (talk) 17:34, 20 June 2015 (UTC)
{{Infobox Country World Championships in Athletics}} done - thanks Plastikspork. Alakzi (talk) 16:00, 25 June 2015 (UTC)
@Alakzi: is this still pending or done? Mdann52 (talk) 18:18, 28 August 2015 (UTC)
China and Japan station are pending. Alakzi (talk) 18:20, 28 August 2015 (UTC)
Hi. I'm working on eliminating the backlog of requests, one request at a time. Unless, someone else takes this one, I will hopefully get to it soon.—cyberpowerChat:Online 00:37, 29 August 2015 (UTC)
@Cyberpower678: Mind if I steal this one, if you haven't started on it yet? I'm working on clearing WP:TFD/H and there's a handful of templates that can be handled in one BRFA, including the two remaining here. ~ RobTalk 16:06, 3 September 2015 (UTC)
Sure. As long as I haven't plastered a doing or coding template, I haven't taken it yet.—cyberpowerChat:Limited Access 16:10, 3 September 2015 (UTC)
Pictogram voting wait.svg Doing... Thanks. ~ RobTalk 16:15, 3 September 2015 (UTC)
Symbol wait.svg BRFA filed ~ RobTalk 04:28, 26 September 2015 (UTC)

Updating US Census Estimates[edit]

Is there a bot available that could add the current United States Census Bureau population estimates (and unfortunately I wouldn't trust OCR for a lot of the older Census files because I often have to look carefully/zoom myself to tell 3 from 8 or 6 from 0)? It should be a fairly straightforward task. The Census updates can be found at I am in the process of adding data (mostly, I am using an AWK script on my computer to format data from a spreadsheet for copy/paste into Wikipedia) manually, and for that, I'm okay, since it gives me a chance to do spot edits on those pages as well and allows me to try to make sure that adding the USCensusPop widget doesn't completely screw up formatting of the page, but it's not something I could do every year.

Specifically, it could check to see if a page for a place has a Template:USCensusPop, and then if so, just update. Very simple. I'd write it myself, but it would be nicer if somebody either has code I can reuse or if they could do it all themselves. Thanks. DemocraticLuntz (talk) 23:30, 21 June 2015 (UTC)

Sounds like a great idea.--BabbaQ (talk) 23:47, 26 September 2015 (UTC)

Convert (certain) external links to HTTPS[edit]

Pictogram voting wait.svg Doing... for reasons and explained here (among others) internet traffic should be encrypted. Recently, Wikimedia decided to use HTTPS by default. Which begs the question why we not also convert external links to HTTPS (wherever this is an option). For instance, one of the most-linked websites on Wikipedia, the Internet Archive, actually encourages HTTPS inbound links ever since 2013, yet most of the external links on Wikipedia to them still use insecure HTTP. Also, all Google services offer HTTPS access, and Google encourages one to use it, but there are still thousands of links to Google Books, Google News, and YouTube with HTTP. Long story short, what I am asking for is a simple search-and-replace bot, to convert:

youtube ... you get the idea.

Is it possible to have this done by a bot? --bender235 (talk) 17:43, 27 June 2015 (UTC)

bender235 does the htto to https econversion for extenal link has consensus? I recall some reactions in the past but I can't find any link to some discussion about it. -- Magioladitis (talk) 17:50, 27 June 2015 (UTC)
We had a discussion on VPP that concluded that we should use protocol-relative links to whichever site supports both HTTP and HTTPS equally. However, since Wikipedia moved to HTTPS by default permanently, protocol relative links make little sense. --bender235 (talk) 18:40, 27 June 2015 (UTC)
JFYI: this is why we do this: “Since the Internet Archive site uses HTTPS by default for its connections, Russian ISPs are unable identify which page is being requested by their users, and thus whether it is the one subject to the new ban.” ISPs can no longer interfere with a site's traffic. All they can do is block the entire domain, which sooner or later will cause public protest. --bender235 (talk) 05:53, 29 June 2015 (UTC)
I think for youtube is better to convert to {{Youtube}}? -- Magioladitis (talk) 17:50, 27 June 2015 (UTC)
Yes, raw Youtube links should be converted to {{Youtube}}, but those inside {{cite web}} or similar templates should just be converted to https. --bender235 (talk) 18:40, 27 June 2015 (UTC)
@Bender235: In early 2014 we discussed using protocol-relative links instead. GoingBatty (talk) 18:35, 27 June 2015 (UTC)
It appears that {{Google books}} uses protocol-relative links while {{YouTube}} and {{Wayback}} use https. GoingBatty (talk) 18:39, 27 June 2015 (UTC)

I pointed out to Bender235 that over and above altering links from "http:" to "https:" that changing from a country specific address (such as to .com may deny access to some people, as sometimes there appears to be a restriction in the access to text in one country but not another. Bender235 wants proof of this, but as I have not kept records of it and I make a lot of edits, I will provide one when I come across it, but in the mean time I see no need to change the country domain along with the connection type.

@PBS: As I have repeatedly explained to you, this claim makes no sense at all. If book X is restricted (by copyright) to be seen in country Y, Google will not determine this by the top-level domain you browse to, but by the IP address you are currently browsing from. This concept is called geo-blocking. Please read the article. --bender235 (talk) 19:18, 6 September 2015 (UTC)

It has been pointed out that this sort of edit can easily mask vandalism (see User talk:Bender235#https), so as it is not a change that needs expediting, that must be weighed in whether this is a suitable candidate for automation (rather than for example adding to to a process like AWB to be done when other more specific changes are made). See also User talk:Bender235#AWB, apparently Bender235's AWB access was removed on by user:Materialscientist on 2 July 2015 (it has not been restored. When discussing this on Bender235's talk page it was suggested by Bender235 that the discussion Wikipedia:Village pump (technical)/Archive 138#HTTPS by default was relevant to this and so should probably be included in this conversation.

-- PBS (talk) 09:50, 13 July 2015 (UTC)

Removal of {{Start date}} from {{Singles}} template[edit]

It has become common practice in album articles to use {{Start date}} in the {{Singles}} add-on to {{Infobox album}}. Per Template:Start date/doc: "This purpose of the {{start date}} template is to return the date (or date-time) that an event or entity started or was created. It also includes duplicate, machine-readable date (or date-time) in the ISO date format (which is hidden by CSS), for use inside other templates (or table rows) which emit microformats. It should only be used once in each such template and should not be used outside such templates." i.e. {{Start date}} should only be used in album articles for the album release date, not single release dates. It would be nice to have a bot to clean this up, as this error is currently in who knows how many articles. Chase (talk | contributions) 16:44, 5 July 2015 (UTC)

While we're at it, the bot that would do this should also remove {{Start date}} from AltDate in {{Episode list}}. nyuszika7h (talk) 19:32, 5 September 2015 (UTC)

DOI bot[edit]

Given a reference in the forms

"<ref>doi:10.[four digits]/*</ref>" "<ref>[four digits]/*</ref>" or "<ref>[four digits]/*</ref>",

the bot should insert the full reference into the article page and into Wikidata. It might be extended to add data to existing references that are, say, missing the date of publication.


HLHJ (talk) 12:28, 8 July 2015 (UTC)

See User talk:Citation bot/Archive1#Replacement citation bot? and the immediately preceding section. --Izno (talk) 13:09, 8 July 2015 (UTC)
That discussion does not appear to be leading to getting a bot to start working on the Cite Doi templates. Abductive (reasoning) 19:13, 8 July 2015 (UTC)
The bot in question preempts the need for doing so (were it turned on). Inserting {{cite journal|doi=value}} and then the bot fills in the other data is what the bot does (or did with {{cite doi}}). --Izno (talk) 19:48, 8 July 2015 (UTC)
As for Wikidata, I'm not sure of your intentions, so you will need to clarify. Regardless, that bot would need to be approved at Wikidata, not here. --Izno (talk) 19:49, 8 July 2015 (UTC)
That sounds good, and would do half my request. I hope it's back soon.
Apologies for the lack of clarity. Wikidata has a data format for journal sources, but there is currently no way to create items from citation templates. See this discussion. There are tools for doing it from a DOI; see the tools section here. It seemed to me that co-ordination between bots working on both might be helpful at avoiding duplicates, etc., but I take your point that separate bots might be easier. HLHJ (talk) 14:40, 16 July 2015 (UTC)
There is consensus at WPMED to replace cite DOI with cite journal on medical articles. Doc James (talk · contribs · email) 15:35, 16 July 2015 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── For the record, we have an user tool that can be used to derive {{Cite journal}} from DOIs. Having a bot that can autoexpand DOIs to full citations would be useful. Maybe one could reuse the {{Cite doi}} template for it; the bot would convert it to a {{Cite journal}}. Jo-Jo Eumerus (talk, contributions) 19:01, 29 August 2015 (UTC)

Redirects to academic journals may lack WPJournals template (class=redirect) in their talk pages[edit]

Would it be possible to check, for each page with Template:WikiProject Academic Journals in its talk page, if its redirects also have Template:WikiProject Academic Journals (class=redirect) in their respective talk pages? Thanks! Fgnievinski (talk) 20:52, 18 July 2015 (UTC)

Symbol wtf vote.svg Idea is not well explainedcyberpowerChat:Online 20:14, 27 August 2015 (UTC)
@C678: sorry, here's an example: Proceedings of the National Academy of Sciences of the United States of America has several redirects [1], some of which are correctly tagged with {{WPJournals|class=redirect}} in their respective talk pages (e.g, Talk:PNAS) others that are either blank or as a redirect to the target's talk page (e.g., Talk:Proc Nat Acad Sci). fgnievinski (talk) 22:47, 28 August 2015 (UTC)
Maybe it's better moving the page to the redirect pages than just create a new page?--Kanashimi (talk) 13:05, 1 October 2015 (UTC)

Using Infobox journal's language field to populate Category:Academic journals by language sub-categories[edit]

Could a bot please inspect values entered in field "language" of Infobox journal? Then possibly populate individual sub-categories of Category:Academic journals by language (as per WP:JWG). Thanks! Fgnievinski (talk) 02:14, 30 July 2015 (UTC)

@Fgnievinski: do you mean check they are valid, or add individual articles about journals to the category? Mdann52 (talk) 16:19, 28 August 2015 (UTC)
@Mdann52: the latter, please; thanks for looking into this. fgnievinski (talk) 19:03, 28 August 2015 (UTC)
@Fgnievinski: in that case, this is beyond my capability, I will leave this for someone else to take a look at. Mdann52 (talk) 19:07, 28 August 2015 (UTC)
@Mdann52: a listing of inconsistencies between category membership and infobox language field would be a great start; then one could manually fix as appropriate. fgnievinski (talk) 19:15, 28 August 2015 (UTC)
Actually, even a reverse listing of transclusions by language field value would be helpful (e.g., English, French, etc.); it doesn't even need to break up sub-values (e.g., "English", "French", and eventual "English and French" would be fine). fgnievinski (talk) 00:12, 10 September 2015 (UTC)

Container category diffusion[edit]

Category:Container categories, by the current definition of the notice box, only allow subcategories, no other pages. If possible, I think it would help the maintenance process if a bot could check container categories for pages, and if found, check if they are already categorized in a subcategory of the container category being checked. If they are, remove them from the container category, referencing the subcategory and WP:SUBCAT in the edit summary. --Slivicon (talk) 14:57, 3 August 2015 (UTC)

I also think, such a bot is required -- Pankaj Jain Capankajsmilyo 13:27, 1 September 2015 (UTC)

One-time error checking help[edit]

I've created new maps for almost every city, village, and township in Ohio, and I'm about halfway done with adding them to their articles. This normally works fine, but I periodically make errors, and it would help if a bot could check all 2200 of these pages for errors. In some cases, I've added the map of one place to the article about a different one; for example, here I copy/pasted one article's map into another article. In other cases, I simply haven't switched all the elements of the description correctly; for example, here I used the correct map, but I made a mistake in the caption, since the city's in Trumbull County, not Mahoning County as suggested by the description.

All maps follow a rigid naming convention: Map of COUNTYNAME County Ohio Highlighting PLACE TYPE.png. "Place" is simply the community or township name, and "Type" is City, Village, or Township (note the capital letter). Likewise, all captions follow the same convention: Location of PLACE in COUNTYNAME County, although "Township" is part of the PLACE section for townships (see the caption for Beaver Township, Mahoning County, Ohio). Given this convention, I expect that the bot can handle the situation easily. I'm imagining the following (collapsed so this request doesn't appear so massive):

Nyttend (talk) 18:12, 18 August 2015 (UTC)

@Nyttend: Though this is a bit old, I assumed it was still needed and went ahead coded it up. Have you done all the replacements yet? BMacZero (talk) 07:02, 8 October 2015 (UTC)
Yes, I finished everything a few days after posting this request. Thanks a lot; I assumed that this had disappeared into the eternally-forgotten-requests archive. Nyttend (talk) 22:17, 8 October 2015 (UTC)

Cleanup of "naked" Google Books?[edit]

Right now there are approximately 2500 pages with "naked" google books entries (defined as containing the string > ). I asked on Wikipedia talk:AutoWikiBrowser whether there was any way to combine AWB and the Wikipedia citation tool for Google Books at to help on cleaning these up, and got a response to ask here. Would this be appropriate for a Bot? It *may* also be appropriate to include <ref>[ text]</ref> cleanup as well, but that would be a later request if the first makes sense.Naraht (talk) 17:05, 24 August 2015 (UTC)

It would have to distinguish between reference and non-reference links, at any rate. IMO, changing the links in this way would be a net improvement of the wiki, at any rate. Jo-Jo Eumerus (talk, contributions) 17:20, 24 August 2015 (UTC)
OP:Absolutely. And given the small (but not non-existant) crossover between those users who would use named refs and those who would put a "naked" google book in a ref, I would be *quite* happy to limit this at the start to something like regexp <ref>[^ <]</ref> Naraht (talk) 18:07, 24 August 2015 (UTC)
@Naraht: If no one else pops up in the meantime, I'll take this on once my current BRFA is concluded, and I have sufficient time to sort this all out. Mdann52 (talk) 08:41, 25 August 2015 (UTC)

If you're going down this line (and it sounds worthwhile) I strongly recommend you set the bot to run slowly, so you give people a chance to notice and then feed back on any errors before they're reproduced on multiple pages that may not be actively Watchlisted. --Dweller (talk) 08:49, 25 August 2015 (UTC)

@Naraht: Creating a appropriate cite tag requires human intelligence and eyeballs to correct mistakes. This is a big no-no for Bots. I see a case for a bot (or report) that lists all pages that have at least one naked books reference and allow people to work the report/backlog by sorting the cite tag out. Hasteur (talk) 12:34, 25 August 2015 (UTC)
Just to note that if I do it, it will be a manual (or at best supervised) run, and won't be ongoing. If you want one that runs constantly, I'm not the person to write that. Mdann52 (talk) 15:02, 25 August 2015 (UTC)
  • Support - I support a bot to be made that fixes these issues.--BabbaQ (talk) 23:41, 26 September 2015 (UTC)

Update the lists at WikiProject Fix common mistakes[edit]

We really need a bot that updates the lists at Wikipedia:WikiProject Fix common mistakes#Log. Right now they are being done manually, which is a very tedious process. Some of the entries have not been updated since November of 2014 and there are a bunch of errors that we haven't added because we can't keep up with the ones we list now.

Note: This was brought up before at Wikipedia:Bot requests/Archive 63#Bot to updated lists at WikiProject Fix common mistakes, where it was marked as resolved and archived, despite the fact that we are still doing this by hand. --Guy Macon (talk) 18:40, 29 August 2015 (UTC)

Monitor and circulate old unanswered questions[edit]

Talkpages have disadvantages, not least that many are unwatched so posting a query on them doesn't always get a response from someone who knows about the subject. But we could greatly reduce this with a bot.

If someone posts on an unwatched talkpage we risk having their query linger unnoticed. Would it be possible to have a bot run lists of open talkpage queries to relevant wikiprojects? With a special list for "talkpage queries on pages not tagged for any wikiproject". I'm sure we could get volunteers to go through the default list and either answer queries or tag those pages for relevant wikiProjects, so as hopefully to bring the query to the attention of someone who could answer it.

It would need to ignore threads marked {{done}}, and ideally the reports to each WikiProject should be colour coded and date sequenced so you could differentiate between queries or discussions that more than one editor had participated in and sections on talkpages with only one editor having posted. For talkpages tagged to multiple wikiprojects it would probably help if we also had a template that marked that section as of interested or otherwise to particular wikiprojects, so someone from WikiProject mountaineering could go through some relevant talkpage threads, answer those they could and tag those that were about glaciation, vulcanology or botany so that the bot would know that while the article on that mountain was tagged to several WikiProjects including mountainerring, that particular thread with a question re the most recent eruption was for WikiProject vulcanology ϢereSpielChequers 11:12, 7 September 2015 (UTC)

It's been my understanding that this is what the project tags accomplish on talk pages. You can click the links on those to find a more general audience for queries. ~ RobTalk 19:41, 7 September 2015 (UTC)
Yes, but this would work the other way round, so anyone visiting a wikiProject page could easily see a list of open questions that are likely to be of interest to their WikiProject. The reason being that newbies post questions on talkpages and sometimes they linger till stale. ϢereSpielChequers 16:20, 8 September 2015 (UTC)

Replace stubs category with stub template[edit]

As of the September 2015 dump, there are over 2700 articles with a stub category. (e.g. \[\[Category:[\w\s]+stubs\]\]) Could someone create a bot that would replace the stub category with the appropriate stub template? (If the stub template already exists on the article, just delete the stub category.)

For example, Antikristos contains Category:Folk dance stubs. The Category:Folk dance stubs page contains {{Stub Category|article=[[folk dance]]|newstub=folk-dance-stub|category=Folk dance}}. Therefore, the bot would:

  1. Delete the stub category
  2. Look at the value in |newstub=
  3. If the article does not contain {{folk-dance-stub}}, add {{folk-dance-stub}} at the bottom of the article

Thanks! GoingBatty (talk) 21:09, 7 September 2015 (UTC)

  • Support -I support this idea.--BabbaQ (talk) 23:42, 26 September 2015 (UTC)
Anyone interested in taking this on? GoingBatty (talk) 17:36, 3 October 2015 (UTC)
A BRFA has been filed here; users are invited to comment. — Earwig talk 19:36, 8 October 2015 (UTC)

Per-day redirects[edit]

Typing a specific date, for example "15 February 2013", into our search box, finds articles with that date in references or maintenance templates, before events which happened on that date. Therefore, for every page like Portal:Current events/2013 February 15, we should have redirects from, at least:

and possibly:

and others. Once the backlog is done, we'd need a maintenance bot to create each new day's set.

We also need to decide what to do for dates were no Portal:Current events page exists (mostly pre 1999). Perhaps redirect to the relevant year? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:02, 8 September 2015 (UTC)

Firstly, mass creating redirects is again something that needs consensus from the community, though personally I don't see much harm in creating a bot for this. Nextly, since my bot already creates the pages, it would be a matter of adding a few lines of code to implement the requested redirects.—cyberpowerChat:Limited Access 16:00, 8 September 2015 (UTC)
  • Good idea, but no need to stop at today, there are plenty of future dates already set, so this could be done for a decade or more to come. As for redirecting any dates not created, the year would be a good default. ϢereSpielChequers 16:16, 8 September 2015 (UTC)
  • Editors have mixed feelings about the Portal namespace and it doesn't have the same requirements as mainspace. I think search results like Special:Search/"3 January 2002" is much better than Portal:Current events/2002 January 3. Creating thousands of cross-namespace redirects should definitely get consensus first. PrimeHunter (talk) 23:01, 8 September 2015 (UTC)

Discussion seems to have stalled. How an we take this forward? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 09:07, 15 September 2015 (UTC)

Redirects away from mainspace has always been the exception and not the rule. If you want to change this with thousands of systematic redirects then I think an RFC is in order with notifications in many places, for example Wikipedia:Village pump (proposals), Wikipedia talk:Redirect, Wikipedia talk:WikiProject Redirect, Portal talk:Current events, Wikipedia talk:Cross-namespace redirects, Wikipedia talk:WikiProject Days of the year. One of these would be a better place for an RFC than here. Without a widely advertised discussion in advance, created redirects risk being mass deleted. Note that Portal:Current events/2013 February 15 is transcluded in mainspace at February 2013#2013 February 15. If we make redirects on dates then I think this would be a much better and less controversial target. PrimeHunter (talk) 13:42, 15 September 2015 (UTC)

Mexican digital television stations[edit]

On September 24, 2015, some Mexican TV stations will become all-digital. This implies a change in callsigns, but most Mexican TV station links are redirects, not articles.

As such, some redirects need to be moved to new locations and references to them changed out. For instance, XHBAB-TV must become XHBAB-TDT. A bot to make these moves would be very helpful, and the code will be vital when more than 600 stations do this on December 31. There are enough references that all of them can be changed and the old -TV suffixes can be removed.

Note that some stations have their own articles, and those will be manually moved and updated.

The stations that are redirects and to be moved are:


I will likely need to make one or two more requests, and then on December 31 we will need to have a massive blitz of some 600 of these, so having reusable code is a must for my sanity. Raymie (tc) 21:37, 8 September 2015 (UTC)

@Raymie: So just to make sure, you want all redirects with the prefixes listed above with the "TV" suffix to be changed to the same prefix with the "TDT" suffix? -24Talk 20:15, 17 September 2015 (UTC)
@Negative24: That would be correct. The actual redirects need to be moved and the links to them need to be modified too. And I'll want to be able to do it again in December with 600+ of them. Raymie (tc) 02:38, 18 September 2015 (UTC)
@Raymie: Alright, I will see what I can do but this would be the first "real" task for User:Bot24 so it might be a tiny bit rough from the beginning. -24Talk 02:43, 18 September 2015 (UTC)

Texas Historical Commission atlas has changed information links[edit]

I'm not knowledgeable enough to know if this is possible to correct with a bot, but there are a considerable amount of articles that are affected by this. These atlas links have been used for NRHP citations, as well as other historical marker citations.

The home for the Texas Historical Commission atlas URL remains the same:

However, once you access information, those links have changed. Whatever is linked to THC as sources in articles are now dead links. I just made a recent change to an article. You can see by the diff how it's been changed. — Maile (talk) 22:25, 8 September 2015 (UTC)

Copied from Wikipedia:Village pump (technical)/Archive 140#Texas Historical Commission atlas has changed information links:
Special:LinkSearch finds 718 links to The count includes all namespaces and cases with multiple links on the same page. There are around 370 different articles. currently says: "Welcome to the new Atlas! The original Atlas, now located at, will eventually be phased out in the coming weeks. Please begin transitioning your use to the new Atlas." The links I examined work if atlas is replaced by atlas1 but it sounds like this is temporary. It would be good to find and update to new atlas url's while the old content can be seen at atlas1 (not all url changes are of the same form). PrimeHunter (talk) 22:40, 8 September 2015 (UTC)

Feasability of using a bot to determine and move large number of pages for updated guideline[edit]

I am proposing a redrafting of the Wikipedia:Naming conventions (ships) guideline, at user:Saberwyn/Proposed ship naming and disambiguation conventions update. One of the major changes follows the outcome of a Request for Comment on the matter of ship article disambiguation. The current form of the proposal is that all ship articles requiring disambiguation will be disambiguated in the form "(yyyy)", where yyyy = year of launch. If the proposed guideline is accepted, over 26,000 ship articles (number determined by transclusions of {{infobox ship career}} - selected to include articles with infoboxes on specific ships, as opposed to class articles) will need to be checked for compliance with the guideline, and if not compliant, moved to a compliant title. Could a bot be used to check the titles of all ship articles (as determined per number above), check that year of launch is the method of disambiguation, and if not, move the article?

The bot will have to:

  • Read the article title to determine if it ends in a parenthical disambiguator (Is the last character in an article title a ")" ? )
    • If not, no action needed by bot... if necessary, humans can move these as they are encountered
  • Check for year of launch (possible options are by reading the date in the "|ship launched=" field of the infobox, or by reading the date in "Category:YYYY ship" if present)
    • If not found or not comprehensible, list for human attention
  • If the final characters of the title do not correspond to the year of launch, move the article to a new title with the year as the parenthical disambiguator
    • If unable to move, list for human attention
  • As it is assumed that the previous title was a valid alternate title, deletion of the old title or updating of article links will not be required

A second bot operation (or both passes if the above is not possible) could generate a list of articles that use a civilian ship prefix (which, under the proposal, will also be depreciated as part of an article title in most cases), so that humans can review article titles and move those necessary to a date-disambiguation title. I'm reluctant to suggest using a bot to move this group, as the prefix may be part of the common name for the subject.

So, theoretically (because the proposal may not pass in this form, or at all), is it possible to create/adapt a bot to do this, how difficult would it be, and what technical problems would have to be surmounted? Any opinions on the appropriateness of the proposed method of disambiguation should be directed to the proposal. -- saberwyn 03:45, 13 September 2015 (UTC)

Symbol merge vote.svg Needs wider discussion given that the proposal is still under development. This is feasible, yes. I would propose a change in the workflow though. It makes more sense to check if the parenthetical disambiguator contains the year launched before you check what that year is in the article; it reduces the work/processing time when an article can be skipped. A bot should just check whether a four digit number is present within a set of parentheses in the title. If so, skip it. If not, then look for the |ship launched= parameter. I wouldn't use Category:YYYY ship, because it's possible mistakes have been made in applying that category (listing year built instead of year launched, for instance). The parameter is more explicit in what the date means, so it's less likely to contain errors. You will need specific consensus on what the disambiguator should be, as this was not covered in the RfC you linked. For instance, should we use (1936) or (launched 1936)? ~ RobTalk 16:25, 13 September 2015 (UTC)
As stated above, the current form of the proposal is for (YYYY) only, but that is subject to change. As for the exact methodology, I'm useless at programming, so any suggestions for improving the hypothetical bot run(s) to make it easier for the programmer/operator would be greatly appreciated. Thanks for the info. -- saberwyn 09:01, 15 September 2015 (UTC)

Bot to undelete 400,000 old IP talk pages[edit]

It has been proposed here that an adminbot be used to restore some 400,000 old IP talk pages that were deleted by user MZMcBride. It is to be noted that about 10,000 such pages have already been restored by user:MusikBot per this BRFA. See the BN discussion and the BRFA for further details. (talk) 19:11, 17 September 2015 (UTC)

What benefit to the project is to be achieved by restoring an extremely large number of pages full of stale warnings? DavidLeighEllis (talk) 23:24, 17 September 2015 (UTC)
The stale warnings would not be restored. The pages would be templated with a notice that older warnings or other discussions can be seen in the page history. The benefit is to primarily make the edit history of the page reviewable by the typical editor who may want to know if new activity from the IP address is related to previous activity from the same IP address. bd2412 T 23:34, 17 September 2015 (UTC)
I'm not convinced of the utility either way. In theory, restoring is beneficial (i.e., ideally they wouldn't have been deleted in the first place), but there's also the practical matter of 400,000 new log entries and 400,000 new edits for new transclusions of {{OW}}. There's also the danger of restoring some problematic history, and half a million pages are far too many for human review. I know this problem was present with the initial restorations, and we can avoid restoring pages with multiple deletions, but with so many pages, there can easily be some awful things lurking that we won't be able to detect. — Earwig talk 00:13, 18 September 2015 (UTC)
I agree. I'd like to see consensus for this before doing this. Unless there's a clear and obvious need for them, they can always be restored immediately, on-demand via WP:UNDELETE if someone truly needs to review ancient templated warnings... and let's be honest, the number of people who need to do that with any regularity is practically zero. --slakrtalk / 01:43, 18 September 2015 (UTC)
I'm also going to have to agree. I'm all for transparency and historical value, but restoring this many pages is a massive undertaking. I don't think we'll get undelete requests for these pages; most people won't even notice or bother to look that prior revisions were deleted. It's true that if we (a) target restoring revisions with summary "deleting stale IP talk page" (or whatever is consistent) and (b) skip any pages with multiple entries in the deletion log, we'll probably bypass any deletions done in response to abuse. Nonetheless, I'm quite frankly not comfortable with the idea without broad consensus. The bot has restored a good 10,000 pages, and from the looks of it there are several thousand more deleted by Tawkerbot strictly under the "stale IP talk page" rationale. I can continue to finish restoring those for you and more, no problem, however from a bot operator standpoint I'm more than content with calling it quits at this point MusikAnimal talk 05:25, 18 September 2015 (UTC)
As noted earlier, all MZM deletioms use exactly the same summary: Old IP talk page. BTW Tawkerbot's deletion log looks empty. (talk) 05:55, 19 September 2015 (UTC)

I think it'd be alright to try get a consensus here, as there's no need to fragment the discussion to more pages. I have left a note at WP:VPR and WP:VPT. (talk) 06:07, 19 September 2015 (UTC)

  • Support: The best reason in support of the undeletion of these pages is that there is no reason why content that does not meet the deletion policy should remain visible only to admins. (talk) 06:07, 19 September 2015 (UTC)
    • What value is there in templated "you shouldn't vandalize Wikipedia" messages that are ten years old? Have you looked through the recent MusikBot restorations? Which of these pages do you think contribute to building an encyclopedia? These pages add a lot of clutter to database dumps and database queries for almost no benefit at all. There's no "content" being hidden here from non-admins or anyone else. --MZMcBride (talk) 15:07, 21 September 2015 (UTC)
  • Happy to support this, provided reasonable care is taken, i.e. only undeleting pages with one log entry, only undeleting pages with the specific rationale of "[[WP:OLDIP|Old IP talk page]]". That said, we need an admin willing to take on the task who has the requisite skill at running a bot (not me) and it would probably be worthwhile to check in with a dev to make sure it they're OK with the performance side of things. Also might be a good idea to ask MZMcBride what he thinks of this, in the case of the initial undeletions, apparently the fact that bd2412 was asking for his own deletions to be undone was relevant. Ideally these deletions should not have happened and I see no real benefit in leaving all these edits to be viewable to admins only. Jenks24 (talk) 06:25, 19 September 2015 (UTC)
  • Weak Oppose. I agree with {{The Earwig}}'s arguments. Moreover, undeleting will cause confusions if the IP owner has changed.
    • One remark: Why again there is a bot request that does not have enough consensus in this page? I though the Vllage Pump was the place for discussions not this page. -- Magioladitis (talk) 06:46, 19 September 2015 (UTC)
  • Instead we could just have a WP:REFUND on request. AFAIK no one has ever asked for it. Graeme Bartlett (talk) 09:19, 19 September 2015 (UTC)
    • What do you think is to be achieved by increasing the work for admins at WP:RFUD, rather than having the bot restore them all in one go? As elaborated by others, there is neglible risk of inappropriate content being restored (since pages with multiple deletion log entries would be skipped). (talk) 12:37, 20 September 2015 (UTC)
  • Comment: If the only objections are that it would flood the logs, then do spread the job out over time. Doing these in bursts of 1000 at a time, with 1 burst per day over 400 days, might be better than doing them all at once. davidwr/(talk)/(contribs) 21:37, 19 September 2015 (UTC)
  • Alternative proposal Instead of undeleting 400,000 pages, add a message to all 400,000 pages saying "Historical versions of this page prior to [date] were deleted. To request un-deletion click here [button to click on goes here]" with a similar message including the date of the last deleted edit in the edit summary, then have another admin-bot undelete pages on request. This way only pages that people care about will be un-deleted. davidwr/(talk)/(contribs) 21:37, 19 September 2015 (UTC)
    • While the notice is probably a good compromise, a second admin bot doing automatic undeletion would be a very bad idea, as there would be no way to determine whether restored edits had inappropriate content on them, such as outing, personal information, or grossly offensive content. It may be better to just say to post at WP:REFUND. I doubt this would overwhelm them; how often is someone going to request an old and unused IP talk page be undeleted? ~ RobTalk 21:46, 19 September 2015 (UTC)
      • I seem to recall at least part of the motivation for deleting the pages was so that new users were not bombarded with irrelevant messages about ten-year old edits as, usually, the IP has been reassigned or is used by different users. I doubt any messages delivered to the current users of these IP addresses would be seen as being of any relevance to them. Some may even be a bit intimidated by them. -- zzuuzz (talk) 21:59, 19 September 2015 (UTC)
    • What do u mean by "pages that people care about"? When nonadmins can't see the content or history of deleted pages, how do they know which pages they care about? Anyway, the process is too much of WP:BURO-creep. (talk) 12:37, 20 September 2015 (UTC)
  • Comment Any pages that are more than several years old that are restored should probably be blanked or replaced by a templated statement saying that old content was removed but that it still exists in the page's edit history. Doing this will prevent the issue of confusing new editors. davidwr/(talk)/(contribs) 05:28, 20 September 2015 (UTC)
    • That seems like a pre-requisite, and it's been mentioned a few times. However I'd suggest that leaving 400,000 new messages, on dynamic IP talk pages for example, will result in confusion. -- zzuuzz (talk) 09:42, 20 September 2015 (UTC)
    • FYI, all the pages that would be restored will have their contents replaced with template {{OW}}. And since it's a bot doing them, the IPs won't see the "You have new messages" box. (talk) 12:42, 20 September 2015 (UTC)
  • I too am opposed to this. Lots of work for no benefit in virtually every case. WP:REFUND indeed is the right place to go; if you want to see the deleted content, I'll happily restore it (assuming no major problems with the page), but mass restoration is a bad idea. Nyttend (talk) 14:52, 20 September 2015 (UTC)
    • Agreed. --MZMcBride (talk) 15:03, 21 September 2015 (UTC)
      • Assuming that restoration can be done with no work beyond clicking a button on a bot that already has its script, and no messages will be created for the IPs, I don't see the problem at all. We have hundreds of thousands of existing IP talk pages with warnings or test posts or whatever, with little rhyme or reason separating those from deleting pages with similar content. We'll never be rid of them all, so we may as well provide some uniformity to their treatment. bd2412 T 17:26, 21 September 2015 (UTC)
        • I'd prefer that we delete the others. Accumulating cruft is bad. --MZMcBride (talk) 17:29, 21 September 2015 (UTC)
  • Symbol merge vote.svg Needs wider discussion This discussion really doesn't belong on this page, either way. It will need wider consensus from the community. ~ RobTalk 04:21, 26 September 2015 (UTC)
  • Rather than restore, I suggest we delete all IP talkpages from IPs that haven't edited or been blocked in the last three years. Such pages have no value as the human behind the IP is unlikely to be the same. But there are drawbacks to keeping such pages. As pointed out above they are offputting to new IP editors; We've recently had RFA opposes based on high percentage of User space edits, restoring 400,000 IP talkpages would make some editors more vulnerable to such RFA opposes, whilst deleting the odd million stale IP pages would reduce many editors userspace edit percentage. When I do newpage patrol I focus on redlinked talkpages as these are newbie edits and the stiki and huggle users pay them little heed. Some will be vandalism, but most goodfaith though not always helpful. The vandals I can forget about because once I've warned them the hugglers and stiki users will be watching their next edits. Deleting a million or so stale IP talkpages would make it easier to spot when new IP editors started to edit from those IPs. Obviously the bot would need to ignore any moved pages. ϢereSpielChequers 20:22, 4 October 2015 (UTC)

Tagging of cat talkpages[edit]

Requesting bot assistance to tag talkpages of all categories and subcategories within Category:Establishments in Rivers State by year and Category:History of Rivers State by period with {{NigeriaProject|Rivers State=yes}}. Thanks. Stanleytux (talk) 11:28, 23 September 2015 (UTC)

Only the category talk pages? Also, generally new tagging should use the standard name of the template, in this case {{WikiProject Nigeria}}, rather than a redirect. Anomie 12:32, 23 September 2015 (UTC)
Just the category talk pages for now, no problem the project template still works this way, no redirect {{WikiProject Nigeria|Rivers State=yes}}. Stanleytux (talk) 13:10, 23 September 2015 (UTC)
Anomie please tag the talkpages of the above categories and all of its subcategories with {{WikiProject Nigeria|Rivers State=yes}}. Thanks. Stanleytux (talk) 13:24, 23 September 2015 (UTC)

Processing Shadows Commons[edit]

Recently I did an edit like this to potentially de-eclipse an image Shadowed at Commons.

It occurred to me that as this could be done for most of the images here Category:Wikipedia files that shadow a file on Wikimedia Commons automatically by a bot.

If bots are allowed to move files, then the entire process can be automated, apart from the eventual F2 deletion.

For All items in Category identifed:- Step 1. - Identify image tagged as {{ShadowsCommons}}. Step 1a. - Ceck to see if it is actually shadowing Step 2. - Rename filename.ext to filename (uploadtimestamp).ext Step 3. - Remove {{ShadowsCommons}} from the renamed file. Step 4. - Replace ALL transclusions and non disscusion page links to the file. Step 5. - Tag the created redirect as F2. Step 6. Repeat until category is empty (or only images remaining are protected generics.)

Sfan00 IMG (talk) 19:20, 23 September 2015 (UTC)

Simply appending the upload timestamp does not strike me as a very useful way to disambiguate images. In many cases, surely there is a better name that a human could come up with. For example – to pick a random image from the category – File:Sleepwalker.jpg would be more appropriately renamed to File:Sleepwalker (comics).jpg or File:Sleepwalker (Marvel).jpg than File:Sleepwalker (20080424).jpg — Earwig talk 20:52, 23 September 2015 (UTC)
(edit conflict) Bots do not have the filemove userright. Although it would be possible to give a bot the "file mover" user right. Currently, there are no bots with this additional flag. Avicennasis @ 20:57, 10 Tishrei 5776 / 20:57, 23 September 2015 (UTC)
True, but I see no reason that we couldn't give a bot +filemover (if people supported a bot doing this). I mean, we have bots with +admin, +templateeditor, +rollback, etc. — Earwig talk 21:14, 23 September 2015 (UTC)
I know. Merely pointing it out. (We even have a bot with +autopatrol for some reason, which is redundant to +bot) Avicennasis @ 21:46, 10 Tishrei 5776 / 21:46, 23 September 2015 (UTC)
Besides the example above, there are other issues, e.g. File:BBC Four.svg. There's no need to rename this file on EnWp - it's essentially a duplicate from commons, so the best case here is to just delete the local copy. Stuff like needs human review. Avicennasis @ 21:46, 10 Tishrei 5776 / 21:46, 23 September 2015 (UTC)
  • Oppose A suitable file name needs to be carefully determined by a file mover. Appending a timestamp is at best confusing, as it suggests to people that the timestamp refers to the time when the picture was created, and should be avoided. In some cases, it may be better to nominate the file for deletion on Wikipedia or Commons. --Stefan2 (talk) 11:11, 24 September 2015 (UTC)

Fix T44616[edit]

A bot should fix all remaining instances of T44616. MZMcBride has already fixed some in December 2012; at the time it was Bug 42616. This will make the moves revertible only by administrators. GeoffreyT2000 (talk) 16:20, 24 September 2015 (UTC)

I'm not clear what the point of this is. What's broken by having an extra newline that we need a bot to fix it? — Earwig talk 19:34, 24 September 2015 (UTC)
Symbol merge vote.svg Needs wider discussion This almost certainly falls afoul of WP:COSMETICBOT, and it would need community consensus prior to being carried out. ~ RobTalk 04:16, 26 September 2015 (UTC)

Day pages[edit]

An adminbot needs to revert pages for days in 2003 and 2004 to their last non-redirect version and move them to their corresponding Portal:Current events page (e.g. January 1, 2003 to Portal:Current events/2003 January 1). If the corresponding Portal page already exists, a history merge will be performed (this is why an adminbot is needed). If the day page has history only as a redirect, it can simply be deleted (another reason for an adminbot). Some users that have previously done such moves are Fram, AnomieBOT (a bot), and Waldir. GeoffreyT2000 (talk) 19:31, 24 September 2015 (UTC)

How are the history merges to be done? I mean, using a bot. In my experience it often requires some subjective judgment to decide which revisions to keep for a clean final result. --Waldir talk 20:06, 24 September 2015 (UTC)

Updating T:TDYK every new day[edit]

Every day, the Template talk:Did you know page is updated by moving the Current nominations level 2 section header to one newer day, and adding a new level 3 section header for articles created/expanded on that day. This task is currently done manually by a human. Examples: September 26, September 25, September 22, September 1.

I think this once-a-day task may be done better by a bot. Note that this is my first bot request, so please notify me if I have made any problems. sstflyer 15:34, 26 September 2015 (UTC)

  • Support - I support this idea. It would be useful.--BabbaQ (talk) 23:44, 26 September 2015 (UTC)

Pictogram voting wait.svg Doing... Seems simple enough. This could run exactly at 00:00 UTC if we want. Happy to implement this, I don't think it will be hard MusikAnimal talk 04:42, 30 September 2015 (UTC)

@Allen3 and Mandarax: Any input on this? MusikAnimal talk 04:43, 30 September 2015 (UTC)
The common case of moving the "Current nominations" header down one day and adding a header for the new day should be straight forward. The hard part is dealing with the exceptions. The most common exception is probably when someone living east of Greenwich adds a header for a new day before 00:00 UTC. When a new date header is added early, the "Current nominations" header is almost never moved. Template talk:Did you know being actively edited by humans also means date headers and their associated comments are mangled upon occasion. --Allen3 talk 09:11, 30 September 2015 (UTC)
Allen3 has covered the key exceptions: the bot will need to know that the new day may already be there, and that there should be eight consecutive days (seven prior days plus the new day) in the "Current nominations" section. The bot should probably be written so as to allow for a different number of days in the "Current nominations" section—it hasn't been that long since we went from five prior days to seven, and it's conceivable that the number could change again in the future. (I would expect such changes to occur years apart.) BlueMoonset (talk) 16:07, 30 September 2015 (UTC)
@Allen3, SSTflyer, and BlueMoonset: Got it. My thoughts: First off, add an editnotice and perhaps some embedded comments saying this process is automated now, so that people don't bother trying to do it by hand. They should simply wait until the bot comes by and does it. If the bot detects a heading for the new day was added, but the "current nominations" header wasn't moved, it will only do the latter, and vice versa. Additionally, the bot will produce a daily report of any errors it encountered when trying to parse the page. I do a similar thing with the WP:PERM pages, see Special:PermaLink/682155886. We could transclude the report at the top of the Template talk:DYK, that only contains errors and otherwise is blank, that way it's easily seen and swiftly fixed. The bot will continue to try to do its task until the errors are fixed, or if it detects the job has already been done by hand. I can have a wiki page specify how many days after the current nominations a heading should exist, so you can change it whenever you need to without my assistance. How does that sound? MusikAnimal talk 16:16, 30 September 2015 (UTC)
If there's a wiki page that specifies the number of days, then it has to be fully protected so that only an admin can change it, much like the queue pages at DYK are protected. This is not a number that should be changed without an RfC being run first. BlueMoonset (talk) 05:28, 2 October 2015 (UTC)
Document-properties.svg Coding... No problemo. I am going to assume my development approach is sane and proceed with coding soon MusikAnimal talk 20:12, 2 October 2015 (UTC)

() Symbol wait.svg BRFA filed Sorry for the delay, got held up with other technical work MusikAnimal talk 01:55, 9 October 2015 (UTC)


I request a bot that places the GOCE tag at articles talk pages that has been through a completed work by the GOCE project. A GOCEbot perhaps.--BabbaQ (talk) 23:38, 26 September 2015 (UTC)

@BabbaQ: Could you elaborate on the procedure? How would the bot know when GOCE has signed off on an article, and what template should it use on the article talk page? Is this currently being done now by hand? MusikAnimal talk 04:50, 30 September 2015 (UTC)
{{GOCE}} is sometimes added by copy editors to indicate that a WP:GOCE copy edit was requested and completed. See the last line of the "Instructions for copy editors" at Wikipedia:WikiProject Guild of Copy Editors/Requests. I'm a big fan of BabbaQ's work, but I don't think that this task has consensus. I suggest that BabbaQ bring it up at at place like Wikipedia talk:WikiProject Guild of Copy Editors. – Jonesey95 (talk) 06:03, 30 September 2015 (UTC)

Alright, let's go with Symbol merge vote.svg Needs wider discussion MusikAnimal talk 16:21, 30 September 2015 (UTC)


I request a bot that places a tag at the talk page with a notice that the article in question has been through the week of being that weeks TAFI selected article. As it is not done all the time today. --BabbaQ (talk) 23:40, 26 September 2015 (UTC)[edit]

Mark all links to as dead. The links are being redirected to a another website. However, they are not marked as 404, or soft 404. That includes links to$1/ etc. which has been deleted retroactively from the archives. Examples:

See more detailed reasoning read this on my blog and FindArticles. (tJosve05a (c) 08:25, 30 September 2015 (UTC) 16,609 mainspace matches. (tJosve05a (c) 08:31, 30 September 2015 (UTC)
Josve05a so that I understand the request correctly (and to help guide the answer) what you're looking for is: For every occurrence where the pattern appears inside a ref block (i.e. regex 'ref>*?*?</ref') append a {{deadlink}} template (with appropriate year/month for categorization) just inside the close of the reference tag. Is this correct? Hasteur (talk) 18:36, 30 September 2015 (UTC)
Yes, unless one {{dead link}} already exists. (tJosve05a (c) 19:12, 30 September 2015 (UTC)
Doesn't just has to be in refs, but in all external links, but if that's to complicated, then the refs are good enough. (tJosve05a (c) 19:13, 30 September 2015 (UTC)

Images tagged for Commons Transfer when the image is already at Commons...[edit]

This Catscan query:-[6]=1&sortby=uploaddate&ext_image_data=1&file_usage_data=1

Is there a way for a BOT to handle this periodically? Namely removing the {{Copy to Wikimedia Commons}} tag, so people aren't confused about what ACTUALLy does need to be reviewed transfered?Sfan00 IMG (talk) 10:43, 2 October 2015 (UTC)

  • Waste of time and unnecessary watchlist clutter. If the file has been tagged with {{subst:ncd}}, it is usually deleted on Wikipedia within a few days. There should be no problem if the "copy to Commons" template remains on the file information page during those few days. --Stefan2 (talk) 11:11, 2 October 2015 (UTC)


A bot needs to remove transclusions of Template:Sort from thousands of pages. The template itself can then be deleted. GeoffreyT2000 (talk) 00:28, 3 October 2015 (UTC)

Looking at it most of the uses are in other templates. It cannot just be removed from them it would break them. But it may only be directly included in a small number, and so could be replaced by in theory inlining the code. I write “in theory” as it would be a massive undertaking to identify and fix them. Here’s one if you want to look at it yourself: {{player2}}.--JohnBlackburnewordsdeeds 01:05, 3 October 2015 (UTC)
@JohnBlackburne: Instead of removing the transclusions, replace them with instances of data-sort-value. GeoffreyT2000 (talk) 01:36, 3 October 2015 (UTC)
If you start with {{Fb team}}, you'll take care of 19,000 of the 64,000 transclusions (including about 90% of the transclusions in other templates). There are probably a few more templates that are used in thousands of articles; the aforementioned {{player2}} has 1,352 transclusions, which will probably take care of 90% of the remaining transclusions in templates. Then you'll be left with articles and a few stray templates. I don't see how this can be done with a bot, but there are some smart people around here. – Jonesey95 (talk) 01:44, 3 October 2015 (UTC)
GeoffreyT2000, do you have idiot-proof instructions for replacing this template? I see the "replace it with data-sort-value" instructions, but that's not enough help for me. Also, can you please provide a link to the discussion that led to the deprecation of this template? Thanks. – Jonesey95 (talk) 02:00, 3 October 2015 (UTC)
Seems to me we should avoid the use of the HTML directly and instead use templates. So even if {{sort}} is deprecated, I would say from my opinion alone there isn't consensus for a mass-replacement. --Izno (talk) 03:44, 3 October 2015 (UTC)
  • Strongly opposed. I'm seeing no discussion at talk or elsewhere, and the doc page's replacement guidelines require more complexity than the simple template. Anyone who starts to replace 64,000 templates with more-complicated coding needs to be blocked immediately to prevent disruption and confusion. Nyttend (talk) 12:22, 9 October 2015 (UTC)
There is no apparent consensus for this task. If one is reached in the future, this task can be resubmitted with clear instructions on how to implement the consensus decision. – Jonesey95 (talk) 13:14, 9 October 2015 (UTC)

Symbol merge vote.svg Needs wider discussion

ShadowsCommons Autotagger[edit]

User:Stefan2 has a query here - which is used to identify files which have the same name on Commons and locally, but which may not be the same media?

Would it be possible for a bot task to run through the list periodically to tag files locally unless the local copy is already tagged using whatever CSD F8 uses, or there is a {{Keep Local}} template on the local copy? Sfan00 IMG (talk) 14:52, 3 October 2015 (UTC)

  • Well, I wrote a bot for this purpose some time ago and tagged a few hundred files (see Wikipedia:Bots/Requests for approval/Stefan2bot), but I found that there were so many cases where it was better to do things on Commons instead (such as correcting someone's uploads of Wikipedia thumbnails to Commons, or nominating files for deletion on Commons), so I thought that it was better to go through the files manually instead. In doing so, I also solve lots of the filename conflicts myself. In the end, someone will need to go through the files manually anyway, and having them in list form doesn't make any great difference. --Stefan2 (talk) 15:11, 3 October 2015 (UTC)

Bot Request[edit]

I am requesting you that please accept my bot request because I am wikipedia editor and now I have created 30 pages, so pleasgive me bot to check there mistakes and correct them. Thankyou.--Productable Khan (talk) 16:42, 9 October 2015 (UTC)

X mark.svgN Not done Read this: Wikipedia:Creating a bot. – Jonesey95 (talk) 17:46, 9 October 2015 (UTC)

Citation overkill[edit]

Wikipedia has a Citation overkill policy, which says that you do not need "more than a couple" citations to back up a single claim. I think a bot could easily recognize when five or more citations are mixed together (like this: [1][2][3][4][5]) and address the problem by adding a "too many citations" tag after the mix of 5 or more citations (like this: [1][2][3][4][5][too many citations]). I believe it would be very helpful because the bot can make more users aware of this policy and it would be less likely to happen in the future.--Proud User (talk) 18:06, 9 October 2015 (UTC)

That is (a) an essay, not a policy, and (b) not as clear-cut as you make out. It's always a judgement call, which bots are notoriously bad at. Relentlessly (talk) 18:12, 9 October 2015 (UTC)

WikiProject Pakistan[edit]

I'd like to request a bot to tag all articles, categories, subcategories, templates under the parent Category:Pakistan with Template:WikiProject Pakistan. It's been a while since bot-assisted WP:PAK tags were added in mass (the last time was in early 2012), and I know that there are hundreds of pages that need tagging. A big thanks and a complementary barnstar await any bot who could take the initiative. Many thanks, Mar4d (talk) 02:49, 10 October 2015 (UTC)