User talk:Citation bot

You may want to increment {{Archive basics}} to |counter= 34 as User talk:Citation bot/Archive 33 is larger than the recommended 150Kb.

Note: See also how to prevent the bot from repeating mistakes. See also Help:CS1, and {{cite xxx}} documentation in general.

Note that the bot's maintainer and assistants (Thing 1 and Thing 2), can go weeks without logging in to Wikipedia. The code is open source and interested parties are invited to assist with the operation and extension of the bot. Before reporting a bug, please note: Addition of DUPLICATE_xxx= to citation templates by this bot is a feature. When there are two identical parameters in a citation template, the bot renames one to DUPLICATE_xxx=. The bot is pointing out the problem with the template. The solution is to choose one of the two parameters and remove the other one, or to convert it to an appropriate parameter. A 503 error means that the bot is overloaded and you should try again later – wait at least an hour.

Please click here to report an error.

Or, for a faster response from the maintainers, submit a pull request with appropriate code fix on GitHub, if you can write the needed code.

Cleanup of date

Status

new bug

Reported by

Headbomb {t · c · p · b} 22:31, 8 June 2022 (UTC)

What should happen

[1]

We can't proceed until

Feedback from maintainers

Will do, if not book, and if no existing |date=, and id year matches any existing |year=. Now to right the code, and unit tests. AManWithNoPlan (talk) 11:35, 13 June 2022 (UTC)

Please make this work only with the |volume= parameter, and not |title=.

A date in the title may be genuinely part of the title, e.g. "Hospital opening delayed to August 2023". --BrownHairedGirl (talk) • (contribs) 14:48, 13 June 2022 (UTC)

I am still not sure how to distinguish dates and Issue numbers/names that are dates. AManWithNoPlan (talk) 18:55, 22 June 2022 (UTC)

Issue is for issue number, not the issue date/issue name.

• Smith, J. (2007). "Title". Magazine. No. Summer 2007. p. 23.

should be converted to

• Smith, J. (Summer 2007). "Title". Magazine. p. 23.{{cite magazine}}: CS1 maint: date and year (link)

Same in cite journal, etc. Headbomb {t · c · p · b} 04:31, 23 June 2022 (UTC)

Please remove |year= if you add |date= in this case. Izno (talk) 05:15, 23 June 2022 (UTC)

It we take time to get these right, so I will program in tests before implementing it.AManWithNoPlan (talk) 13:17, 24 June 2022 (UTC)

Mislabeling Associated Press and Reuters as a "work" rather than an "agency"

Status

new bug

Reported by

Dawnseeker2000 22:34, 25 June 2022 (UTC)

What happens

[2] Here, both AP and Reuters are changed from using the "agency" parameter to the "work" parameter.

1. Usage of the "work" parameter is limited to news sources that are websites, newspapers, journals, or magazines. Agencies follow the formatting style for publishers.
2. From the initial sentence of our Reuters article: "Reuters (/ˈrɔɪtərz/ (listen)) is an international news agency". See also that our article title is not italicized. That is correct and consistent with ref formatting for news agencies.
3. From the initial sentence of our Associated Press article: The Associated Press (AP) is an American non-profit news agency. See also that our article title is not italicized. That is correct and consistent with ref formatting for news agencies.

What should happen

Recommend that Associated Press and Reuters be switched to use the "Agency" parameter.

We can't proceed until

Feedback from maintainers

Not a bug. See the template documentation.

—Trappist the monk (talk) 23:03, 25 June 2022 (UTC)

What TMK is referring to is the statement "Do not use for sources published on the agency's own website; e.g. apnews.com or reuters.com; instead, use work or publisher". Using the work parameter presents in italics while publisher does not. I wonder why the indifference. I suggest we use the publisher parameter (again, for consistency with "agency" styling). Dawnseeker2000 23:25, 25 June 2022 (UTC)

I agree with Trappist. In these cases, the news agency is acting as a publisher rather than as an agency. Neutral on whether to use |work= or |publisher=. --BrownHairedGirl (talk) • (contribs) 13:03, 26 June 2022 (UTC)

When the publisher and work are the same, publisher is not usually used. AManWithNoPlan (talk) 14:29, 26 June 2022 (UTC)

Associated Press and Reuters are not works though. They are agencies/publishers, and should not be converted to works. Headbomb {t · c · p · b} 15:09, 26 June 2022 (UTC)

Review the template documentation. Realize accordingly that your comment here is in direct contradiction to your comments above about Cite magazine. Yes, these are the same exact issue. Izno (talk) 16:40, 26 June 2022 (UTC)

If this was directed at me... which contradiction? Dawnseeker2000 00:03, 27 June 2022 (UTC)

AP and Reuters are both news agencies, and news works in their own right. When citing the AP or Reuters website directly they are works. When citing a story on another news organisation's website that says that the reporting is from AP or Reuters, then they are acting as an agency. Sideswipe9th (talk) 17:06, 26 June 2022 (UTC)

When used as the name of a work (for an article directly from the AP web site), it should have |work=AP News or |work=Associated Press News. That is the name they give to that part of their site, that is, the work. It is incorrect to list |work=AP or |work=Associated Press. That is the name of the organization, not the name of their web site, and should appear in |publisher= or |agency= instead. —David Eppstein (talk) 19:51, 26 June 2022 (UTC)

Yes, that's what I'm saying. News outlets (regardless of media type) are considered publishers Dawnseeker2000 01:18, 27 June 2022 (UTC)

RfC: Should you use cite web, or cite magazine, or cite news?

@Sideswipe9th: Please don't hold RfCs in user talk space (the practice of holding RfCs to discuss user conduct ceased some years ago). This matter should be discussed at Help talk:Citation Style 1. --Redrose64 🌹 (talk) 21:29, 1 July 2022 (UTC)

@Redrose64: I had asked previously where to hold this and there was an even split between this talk page, and the CS1 talk page being the appropriate venue. I'm not opposed to moving it, if it was to be relaunched, so to speak, with the same question, would copy and pasting what has already been discussed and !voted on be acceptable?

I'm also not opposed to using this opportunity to address the concerns raised by both @XOR'easter and David Eppstein:, which may result in a different question being asked however. I'd just like to check what the options are. Sideswipe9th (talk) 21:41, 1 July 2022 (UTC)

I've moved the whole RfC unchanged, apart from a slight adjustment to the section heading and also omitting the two posts above. --Redrose64 🌹 (talk) 08:14, 2 July 2022 (UTC)

Moved to Help talk:Citation Style 1 § RfC: Should Citation bot use cite web, or cite magazine, or cite news?

 – Redrose64 🌹 (talk) 08:14, 2 July 2022 (UTC)

DOING MORE HARM THAN GOOD! Stop this bot!

Status

new bug

Reported by

Arminden (talk) 06:17, 24 July 2022 (UTC)

What happens

bot removing essential page number and replacing it with incorrect or insufficient ones

What should happen

deactivate the bot

Relevant diffs/links

https://en.wikipedia.org/w/index.php?title=Hezekiah_(governor)&diff=next&oldid=1044424095

We can't proceed until

Feedback from maintainers

Your bot MUST be removed! You are damaging articles with no real benefit to balance that.

At Hezekiah (governor), this edit added data of little importance (doi, jstor, s2cid) since the URL is already indicated, while removing the very much essential page number and replacing it with incorrect or insufficient ones:

• "page= 122–126 [125]" is correct data, in both form and content. It means that the entire article covers pp. 122–126, while the relevant ref page is 125 only.
• "pages= 8, 13" are the correct, relevant ref pages. The bot replaced them with who knows what ("109–118"), maybe the general pages of the article.

What really matters is the actual page the ref is citing. Some sources can't be fully accessed online, but the ref page can be. If one can indicate the start & finish pages, that can benefit some who do have access (subscribers, access to hard copies), but is not essential. The form "x-y [z]" is well accepted in academic publications and transparent enough for the user: "the article goes from x to y, but look up only z." I have also seen it as "x-y (z)". Now you have forced me to remove the start & end pages of the article, in order to preserve the relevant page number(s) I'm actually citing. What's far worse, elsewhere the intervention of the bot has most likely not been reverted and now the cited page is missing. Now you must go and manually fix all the damage already done.

The bot harms the user. We're far better off without it. Arminden (talk) 06:17, 24 July 2022 (UTC)

|page=122–126 [125] should be kept when converting to the proper |pages=, yes. But |pages=8, 13 is definitely incorrect. Those are the pages of the preprint PDF, not the pages of the published book, which is what the reader [thus bot] is expecting.

Also, you were not harmed by the bot. Don't be dramatic. Headbomb {t · c · p · b} 06:52, 24 July 2022 (UTC)

The template is problematic, as there's only one page of reference, 125, so definitely singular, page. But it can be useful to indicate the size and position of the article, 122–126. I'm not coding, those who are should find a solution.

As I wrote on your page: your manual edits are absolutely perfect, detailed, and most welcome; the bot however is not ready to be let go. Unless and until it is, it is totally wrong to activate it.

"Being dramatic" is a touch too personal, if you don't mind. Wiki IS better off without the unfixed bot, broken bots tend to damage lots of articles, and those in charge aren't always quick to react, so clear words can only help. Arminden (talk) 10:45, 24 July 2022 (UTC)

fixed- now detects square brackets or commas AManWithNoPlan (talk) 11:37, 24 July 2022 (UTC)

AManWithNoPlan, great, thanks! Now that I know you can do it, let's see all aspects.

• "page= X–Y [Z]" means that only Z is directly relevant, so "page" in the singular. Or not? Ideally the template should lead to smth like "pp. X-Y (see p. Z)", or if needed "pp. X-Y (see pp. V, W, Z)".
• "page= X–Y (Z)" is also used by some. I also seem to remember that in some templates or contexts, straight brackets create a lot of chaos, so round ones are an important option. Arminden (talk) 11:56, 24 July 2022 (UTC)

PS: All the edits produced by the bot before it reaching the final, mature and correct form must be fixed retroactively. Does that happen automatically, or must they be identified and fixed by hand, one by one? If the latter, how and who? Arminden (talk) 12:00, 24 July 2022 (UTC)

round paras add too. The bot has been fully approved for well over a decade. People are expected to report problems as they see them. AManWithNoPlan (talk) 12:10, 24 July 2022 (UTC)

Why on Earth would the reader be expecting the pages of the published book and not the linked copy? On the various templates we provide links to preprints, partial google books previews, and chapter excerpts, in spite of the bulk of the citation indicating the full published work, entirely because we fully expect the bulk of the readers are not going to have journal subscriptions or a complete next-door library and thus, should they want to verify a statement, they should be given as direct a route as possible.

Changing page numbers in any form is absolutely uncalled for unless you, personally, verify they still point to the sourced information. SamuelRiv (talk) 12:17, 24 July 2022 (UTC)

""Why on Earth would the reader be expecting the pages of the published book and not the linked copy?" because that is the published version. I'm the one that added the preprint link. It wasn't linked before. If I tell you look up something from Hoogenboom, BJ; Manske, RC (October 2012), "How to write a scientific article", International Journal of Sports Physical Therapy, 7 (5): 512–7, PMID 23091783, you're expected to look at the published version of that, not a preprint of it. Headbomb {t · c · p · b} 13:52, 24 July 2022 (UTC)

SamuelRiv, hi. We are almost of the same opinion, but just almost. Indicating the pages of the actually linked PDF (a preprint) is what we had and what's most needed, the bot changed that, and then the editor who apparently activated it, Headbomb, set in both infos.

Why the full info can be of some use? Because links rot, people do have access to hard or digital copies, and I for instance often bump into JSTOR docs I can only read online and knowing if the needed page is not too far in (not, say, on the 20th page of the doc) helps me decide right away if to bother clicking through the doc. I do read some of the referenced sources, as they educate me about topics I'm interested in, and often bump into "failed verification" cases. For some important titles I do have & use hard copies. So it's not all that hypothetical. ANd if it does no harm by replacing potentially useful, valid data with bad, let it be, it's a bot, doesn't cause sweat.

What I didn't realise: apparently, the bot isn't automatically crawling through the entire enWiki and doing edits, but needs to be activated by a specific editor for each article. Or not? Very different concepts. Arminden (talk) 13:27, 24 July 2022 (UTC)

Both modes are authorized. Generally a user has a specific issue that needs fixe, such as all pages where the jstor url has #meta_blah_blah_blah included, or all pages with a certain invalid DOI, etc. AManWithNoPlan (talk) 13:42, 24 July 2022 (UTC)

In the case where both the page range and pinpoint page(s) are given, the pinpoint should be preferred as the page range is almost always easily determined with the rest of the citation information, and the pinpoint is far more important for verifiability. If the bot can't distinguish in a template whether the editor was referencing page range, pinpoint, both, or neither, the proper action is to leave it alone or, if the metadata are corrupted, convert to |at=. SamuelRiv (talk) 18:06, 27 July 2022 (UTC)

World Bird List

The bot has repeatedly changed the "cite web" to "cite journal" in this reference though it is clearly a web page and not a journal (<> and {} brackets removed so it shows here): cite web ref name=IOC12.1Gill, F.; Donsker, D.; Rasmussen, P. (July 2021). "IOC World Bird List (v 12.1)". doi:10.14344/IOC.ML.11.2. Retrieved January 15, 2022. /ref . Craigthebirder (talk) 17:45, 27 July 2022 (UTC)

Not only in bird lists, in individual species' pages. Craigthebirder (talk) 18:52, 27 July 2022 (UTC)

Why are you citing version 12.1 in the title and version 11.2 in the doi (which doesn't actually link to version 11.2 of the list)? And for specific bird species why aren't you linking to the particular place in the list for that species – shouldn't you link to https://www.worldbirdnames.org/new/bow/buttonquail/ from Plover?

—Trappist the monk (talk) 19:01, 27 July 2022 (UTC)

First question - because I messed up; have corrected for future and will try to correct earlier entries. Second question - because that page is the link to the whole species list. But will link to family sections in the future. Thanks for pointing these out. Craigthebirder (talk) 20:34, 27 July 2022 (UTC)

I flagged 10.14344 as a non-journal DOI. AManWithNoPlan (talk) 20:01, 27 July 2022 (UTC)

Thank you. Craigthebirder (talk) 20:34, 27 July 2022 (UTC)

CS1 errors

This bot routinely causes CS1 errors - for example, [this edit] caused the article to be listed at https://en.wikipedia.org/wiki/Category:CS1_maint:_numeric_names:_authors_list. It happens often, making unnecessary busywork to those of us who do CS1 maintenance. The bot should be modified so that it doesn't make any CS1 errors and cause pages to be listed at either https://en.wikipedia.org/wiki/Category:CS1_maintenance or https://en.wikipedia.org/wiki/Category:CS1_errors

Ira Leviton (talk) 20:37, 28 July 2022 (UTC)

Bad metadata is an ongoing issue yes. There's no way to anticipate all cases. Headbomb {t · c · p · b} 21:37, 28 July 2022 (UTC)

As with any other script, users of Citation bot are responsible for verifying that edits made using that feature are correct. --Redrose64 🌹 (talk) 22:46, 28 July 2022 (UTC)

New York added to non-human list. AManWithNoPlan (talk) 17:33, 31 July 2022 (UTC)

Untitled_new_bug

Status

new bug

Reported by

SteveMcCluskey (talk) 16:40, 31 July 2022 (UTC)

Relevant diffs/links

https://en.wikipedia.org/w/index.php?title=Aberration_(astronomy)&diff=1101370619&oldid=1101370242

We can't proceed until

Feedback from maintainers

The edit at Aberration_(astronomy) changed the pages= entry to the issue number (e.g., A91), removing the page range (e.g. 1-6) of the article. This happened several times. I recently corrected the pages= entries.

You're actually reintroducing errors the bot is fixing (full proper fix is here, there were some GIGO issues). A&A and similar journals use an 'article ID' instead of a page number. This is what goes in |page= and how people cite these journals. There are no issue numbers for those. Headbomb {t · c · p · b} 17:07, 31 July 2022 (UTC)

I'm puzzled by this usage. There are clearly page numbers in the pdf version of the article; where should they be provided if not in the page(s)= entry of the citation template. The article ID seems analogous to an issue number. Could you point me to appropriate reference supporting this strange usage of the citation template. --SteveMcCluskey (talk) 18:08, 31 July 2022 (UTC)

OK, I see that you're using the page= value to place the ArticleID without parentheses, which are used for issue numbers. It's a bit of a kludge but it reproduces the style used in AandA. However, it leaves no place for the page numbers (which this historian of science finds standard for a complete citation). I'm not going to make a big deal of it. --SteveMcCluskey (talk) 18:31, 31 July 2022 (UTC)