Help talk:Citation Style 1

|page= same value as last n-digits of |doi=

I'm seeing more and more of this sort of misuse of |page(s)=. What to do about templates like this where |page(s)= matches the trailing digits of |doi=?

{{cite journal |last1=Kesari |first1=Vigya |last2=Ramesh |first2=Aadi Moolam |last3=Rangan |first3=Latha |title=''Rhizobium pongamiae'' sp. nov. from root nodules of ''Pongamia pinnata'' |journal=BioMed Research International |volume=2013 |pages=165198 |year=2013 |pmid=24078904 |pmc=3783817 |doi=10.1155/2013/165198 |doi-access=free }}

Example template taken from this edit.

If we look at the pdf version of the article, we can see that 165198 is the article ID so 165198 properly belongs in |article-number=.

Maintenance message? {{citation}} with |journal= and {{cite journal}} templates only? Value assigned to |page(s)= must be greater than some threshold value, perhaps 9999? Other constraints?

—Trappist the monk (talk) 16:44, 18 April 2025 (UTC)

The example came from this edit[1] a few years prior. I thought this would be citoid being flaky, but it just ignores the article number (at least now it does).
I've come across a limited selection of journals that run there page numbers across issues in the same year, but even then they don't get to over 10k page numbers. A maintenance message would be good as the page number is obviously invalid, although I wouldn't think there's that many. -- LCU ActivelyDisinterested «@» °∆t° 20:22, 18 April 2025 (UTC)

This crude search (times out) suggests that there are more than 4000 articles that have |page(s)= with five-digit values. There are journals that do use 5 digit pagination: doi:10.1073/pnas.1203495109

—Trappist the monk (talk) 20:51, 18 April 2025 (UTC)

As you know my ability to write regex extends to asking someone else how to do it, but I expected that that would time out. Could the code be set to check configuration and ignore certain journals? It would allow a method to hide false positives en masse, 10.1073 alone has tens of thousands of uses. -- LCU ActivelyDisinterested «@» °∆t° 23:22, 18 April 2025 (UTC)

Could the code be set to check configuration and ignore certain journals? What code are you asking about? The regex? The cs1|2 module? Sommat else?

If the regex: not really. We can exclude all articles that have the doi prefix 10.1073 (~26000). This search found ~3350 articles but it also times out.

If the cs1|2 module: yes if necessary.

—Trappist the monk (talk) 13:32, 19 April 2025 (UTC)

Sorry that was ambiguous, I did mean the cs1|2 module. That way any other journals that commonly use such large pages numbers could be excluded using the same method. -- LCU ActivelyDisinterested «@» °∆t° 18:10, 19 April 2025 (UTC)

I'm not convinced that we need to filter by doi prefix. Generally, these days, {{cite journal}} templates use |page(s)= to list a page-range so a journal citation with five-digit pagination will contain some form of dash. When a |page(s)= value has dashes, commas, or semicolons, the code will skip the test.

Journals use a variety of formats for article numbers: some are all digits, some are alphanumeric, some begin with a lowercase 'e', some numeric forms include a dot in the matching doi. It may not be possible to create a test that will find all possible article numbers in |page(s)= but we can start with these more-or-less common formats and see where that gets us.

—Trappist the monk (talk) 19:02, 19 April 2025 (UTC)

Physical Review journals have pages in the 6 digit range, e.g. doi:10.1103/PhysRevLett.89.170405. 22:41, 19 April 2025 (UTC) Headbomb {t · c · p · b} 22:41, 19 April 2025 (UTC)

But those seem to be article numbers, not page numbers. Looking at the PDF for the article you linked, I see pages numbered from "170405-1" to "170405-4". FWIW, what I assume to be the CCC string is "0031-9007=02=89(17)=170405(4)$20.00".

Solomon Ucko (talk) 21:22, 27 April 2025 (UTC)

As an experiment, I have written an awb script that trawls articles looking for {{cite journal}} templates that have |doi= and |page(s)=. The list of tested articles is more-or-less the first 25000 articles that transclude {{cite journal}} – the same list you would get from Special:WhatLinksHere/Template:Cite journal. The most recent test found 9575 {{cite journal}} templates that have |doi= and |page(s)= that meet certain criteria. The script skips templates that have |page(s)= values that:

• are ranges separated by underscore, hyphen, emdash, endash, figure dash, or minus character
• are comma- or semicolon-separated lists of pages
• have external urls
• are digit-only values less than 10000

For those templates that are not skipped, the script compares the trailing (rightmost) characters of the |doi= value against the whole value of the |page(s)= value. When there is a match, the script saves the |doi= and |page(s)= values to a file with the test result.

When the first test does not indicate a match, the script modifies the |page(s)= value:

• when the value is exactly 8 digits, the script inserts a dot between the fourth and fifth digits and retests against the |doi= value
• when the first character is a lowercase e, the script removes that character and retests against the |doi= value
• when the first two characters are uppercase CD, the script retests against the |doi= value
• when the first two characters are uppercase CD, the script retests against the |doi= value but accepts values where the |page(s)= value is suffixed by .pubn where n is any single digit

With those simple tests, the script identified 8611 templates that likely are using |page(s)= when they ought to be using |article-number=.

• 3604 numbers only (example)
• 319 dotted 8-digit numbers (example)
• 687 alpha-numeric (includes some CD and e prefixes) (example)
• 1199 CD prefixes with .pubn suffixes (example)
• 2802 e prefixes (example)

The quantities mentioned reflect the number of templates with these various characteristics, not the number of articles.

—Trappist the monk (talk) 15:19, 22 April 2025 (UTC)

8k is small enough that it could be cleared down without any of the things I mentioned. -- LCU ActivelyDisinterested «@» °∆t° 11:17, 25 April 2025 (UTC)

The 8k number is the number of templates found in 25k articles. {{cite journal}} is used in approximately 1.1 million articles so we can expect perhaps something like (1,100,000/25,000) * 8,000 = 352,000 articles in the maintenance category.

—Trappist the monk (talk) 13:19, 25 April 2025 (UTC)

I have hacked Module:Citation/CS1/sandbox to implement prospective detection:

• {{cite journal/new |title= Title |journal=Journal |doi=10.1016/j.gfs.2020.100393 |page=100393}} – |page(s)= is five or more digits

  "Title". Journal: 100393. doi:10.1016/j.gfs.2020.100393.{{cite journal}}: CS1 maint: article number as page number (link)

• {{cite journal/new |title= Title |journal=Journal |doi=10.1002/14651858.CD004052|page=CD004052}} – |page(s)= is five or more characters

  "Title". Journal: CD004052. doi:10.1002/14651858.CD004052.{{cite journal}}: CS1 maint: article number as page number (link)

• {{cite journal/new |title= Title |journal=Journal |doi=10.1126/sciadv.adl0822 |page=eadl0822}} – |page(s)= is five or more characters with e prefix (case insensitive)

  "Title". Journal: eadl0822. doi:10.1126/sciadv.adl0822.{{cite journal}}: CS1 maint: article number as page number (link)

• {{cite journal/new |title= Title |journal=Journal |doi=10.1098/rsbl.2017.0301|page=20170301}} – special case: |page(s)= is exactly eight digits modified by the test to insert a dot between the 4th and 5th digits

  "Title". Journal: 20170301. doi:10.1098/rsbl.2017.0301.{{cite journal}}: CS1 maint: article number as page number (link)

• {{cite journal/new |title= Title |journal=Journal |doi=10.1002/14651858.CD005216.pub2|page=CD005216}} – special case: |page(s)= has CD prefix (case insensitive) followed by typically six digits modified by the test to append .pubn suffix where n is a single digit

  "Title". Journal: CD005216. doi:10.1002/14651858.CD005216.pub2.{{cite journal}}: CS1 maint: article number as page number (link)

Needs a better category name; page will most likely be read to mean 'article' which is not the intended meaning. Suggestions?

—Trappist the monk (talk) 22:27, 27 April 2025 (UTC)

Changed category name from 'CS1 maint: page has article number' to 'CS1 maint: article number as page number' which is a smidge better, perhaps?

—Trappist the monk (talk) 22:03, 10 May 2025 (UTC)

TM:Cite FTP

This does not exist but probably should.

FTP is a long story. Prior to 2020/2021, FTP links were available through a browser at ftp:// and it worked thus we had many {{cite web}} with ftp:// URIs. After that, most modern web browsers stopped supporting the FTP protocol, due to security concerns. Some sites have HTTPS gateways so you only needed to change ftp:// to https:// and it works:

• ftp://naif.jpl.nasa.gov/pub/naif/generic_kernels/spk/planets/a_old_versions/de418_announcement.pdf
• https://naif.jpl.nasa.gov/pub/naif/generic_kernels/spk/planets/a_old_versions/de418_announcement.pdf

I did a bot run in 2021, detected when a gateway existed, and converted the URI from ftp:// to https://

The problem is ftp:// links that have no https:// gateway like ftp://ssd.jpl.nasa.gov/pub/eph/planets/ .. They are only accessible via a dedicated FTP client eg. the unix ftp command. These URIs do not work with web browsers, without special and possibly insecure add-ons. There are a couple issues for us:

1. Semantically they are not on the web, and thus {{cite web}} is inaccurate
2. Web browsers by default do not work and we should not be providing hyperlinks, if nothing else for security reasons.

Solution: We need a FTP template that will display the FTP URI without hyperlinking it. It will document how to access ftp:// links, such as through a user-supplied FTP program. It would also encourage the reader to find an alternative method - a https:// at a different location, or a FTP-https gateway - and encourage replacing the ftp:// with a https:// where possible.

-- GreenC 19:19, 27 April 2025 (UTC)

FWIW, even if the browser itself can't usually handle FTP nowadays, it can still open another application that supports the scheme. For example, clicking the FTP link on my phone asks me if I want to open it in a file manager app I've installed that supports FTP, which works. So having the clickable link can still be useful sometimes.

Therefore, what would make sense to me would be to keep the existing link, but also add some kind of note or explanation, perhaps linking to a page with instructions for accessing FTP on various operating systems, as well as recommending first trying SFTP and FTPS in hopes that the server supports either of these secure variants.

Does FTP have any vulnerabilities beyond those of HTTP without SSL/TLS? What risks could using add-ons realistically pose?

Solomon Ucko (talk) 20:48, 27 April 2025 (UTC)

I don't know that {{cite FTP}} as a native cs1|2 template is something we should do. You could create a wrapper template around {{citation}} using Module:template wrapper. The wrapper would make sure that |url= has a value using the ftp:// scheme (error message if some other scheme); might append the nowiki'd url to the value in |title= or might force |type=FTP if the link is clickable; might provide a link to the explanatory text that you mentioned. Such templates might look like this (using {{citation}} as a mockup):

{{cite FTP |author=W. M. Folkner |author2=E. M. Standish |author3=J. G. Williams |author4=D. H. Boggs |date=2 August 2007 |title=Planetary and lunar ephemeris DE418 |work=Jet Propulsion Laboratory |url=ftp://naif.jpl.nasa.gov/pub/naif/generic_kernels/spk/planets/a_old_versions/de418_announcement.pdf |link=no}}

W. M. Folkner; E. M. Standish; J. G. Williams; D. H. Boggs (2 August 2007). "Planetary and lunar ephemeris DE418 ftp://naif.jpl.nasa.gov/pub/naif/generic_kernels/spk/planets/a_old_versions/de418_announcement.pdf". Jet Propulsion Laboratory. – For help viewing ftp documents see <link>

{{cite FTP |author=W. M. Folkner |author2=E. M. Standish |author3=J. G. Williams |author4=D. H. Boggs |date=2 August 2007 |title=Planetary and lunar ephemeris DE418 |work=Jet Propulsion Laboratory |url=ftp://naif.jpl.nasa.gov/pub/naif/generic_kernels/spk/planets/a_old_versions/de418_announcement.pdf |link=yes}}

W. M. Folkner; E. M. Standish; J. G. Williams; D. H. Boggs (2 August 2007). "Planetary and lunar ephemeris DE418" (PDF). Jet Propulsion Laboratory (FTP). – For help viewing ftp documents see <link>

I guess I prefer the second option because it's cleaner and it doesn't hurt to have a live link.

—Trappist the monk (talk) 00:30, 28 April 2025 (UTC)

I'm OK with this proposal. Sadly I don't have the time or experience to make a wrapper template. I would though be able to populate it via bot. -- GreenC 06:05, 1 May 2025 (UTC)

Why use a separate template? The CS1 templates can have "doi:" at the end, so why treat the FTP address so differently? Rjjⁱⁱⁱ (talk) 00:33, 28 April 2025 (UTC)

Possibly. It could be messy to adapt CS1 to the requirements: special documentation and help pages, gateway vs. non-gateway, display vs. non-display of URIs. I don't think we are talking a ton of URIs either. Mostly the sites are legacy government or university that are too cheap to change and don't care about data security (climate data etc). When I did the conversion bot in 2021 I found tons of old FTP links that no longer work, every year there are fewer. Isolating them to their own template makes it easier to manage the corpus. -- GreenC 06:33, 1 May 2025 (UTC)

@GreenC, I have started a wrapper template, based on discussion above and several published style guides:[2][3][4][5] No docs yet. It'll do a preview warning and maintenance category with no URL. It works almost just like {{citation}}. It gives CS1 formatting by default. Anything (no, none, nope, n) in the |link= parameter suppresses it right now. Feel free to change anything that I have wrong in the template; I'm just getting a ball rolling.

Default

{{cite FTP |author=W. M. Folkner |author2=E. M. Standish |author3=J. G. Williams |author4=D. H. Boggs |date=2 August 2007 |title=Planetary and lunar ephemeris DE418 |work=Jet Propulsion Laboratory |url=ftp://naif.jpl.nasa.gov/pub/naif/generic_kernels/spk/planets/a_old_versions/de418_announcement.pdf}}

W. M. Folkner; E. M. Standish; J. G. Williams; D. H. Boggs (2 August 2007). "Planetary and lunar ephemeris DE418" (PDF) (FTP). (To view documents see Help:FTP.)

No link

{{cite FTP |author=W. M. Folkner |author2=E. M. Standish |author3=J. G. Williams |author4=D. H. Boggs |date=2 August 2007 |title=Planetary and lunar ephemeris DE418 |work=Jet Propulsion Laboratory |url=ftp://naif.jpl.nasa.gov/pub/naif/generic_kernels/spk/planets/a_old_versions/de418_announcement.pdf |link=no}}

W. M. Folkner; E. M. Standish; J. G. Williams; D. H. Boggs (2 August 2007). "Planetary and lunar ephemeris DE418" (PDF) (FTP). {{cite web}}: Unknown parameter |link= ignored (help) (To view documents see Help:FTP.)

With access-date

{{cite FTP |author=W. M. Folkner |author2=E. M. Standish |author3=J. G. Williams |author4=D. H. Boggs |date=2 August 2007 |title=Planetary and lunar ephemeris DE418 |work=Jet Propulsion Laboratory |url=ftp://naif.jpl.nasa.gov/pub/naif/generic_kernels/spk/planets/a_old_versions/de418_announcement.pdf |access-date=3 May 2025}}

W. M. Folkner; E. M. Standish; J. G. Williams; D. H. Boggs (2 August 2007). "Planetary and lunar ephemeris DE418" (PDF) (FTP). Retrieved 3 May 2025. (To view documents see Help:FTP.)

Rjjⁱⁱⁱ (talk) 04:21, 4 May 2025 (UTC)

This is wonderful User:Rjjiii very happy with this. Curiously the document "How to Access an FTP Server from the Browser" proceeds to say it can't be done, and to use a dedicated FTP application. We could have an in-house wiki page in the help: namespace. Like Help:FTP similar to Help:Archiving a source, then link the external "How to" page from there. And link to the Help: page from the template.

FTP is in 5,136 pages. Most are {{cite web}} and have no https interface. I think conversion will be simple: change the name of the template, and change |website= to |work=. User:Trappist the monk, is that a reasonable plan, are there any other parameters cite web has citation does not? -- GreenC 06:18, 4 May 2025 (UTC)

I've come to believe that the |link= parameter should not be supported. A convenient link is far superior to an unlinked string of url gibberish – cs1|2 had a reason for linking |title= with |url=: this is an electronic encyclopedia; all of those published style guides are for works that have physical form. So, I see no reason for {{cite FTP}} to exclude |url=, |url-access=, |url-status=, |archive-url=, |archive-format=, and |access-date= or, frankly, any of the other parameters that are currently excluded: |chapter-url=, |contribution-url=, |section-url=, |article-url=, |entry-url=, |map-url=, |section-url= – {{citation}} does not support |map-url= or its companion |map=. The other excluded url-holding parameters also have companion 'title-holding' parameters. If we are to exclude all of these url-holding parameters, ought we not exclude their companions? Further, placing the ftp url at the end of the rendered citation prevents it from being included in the citation's metadata.

Since |link= should go away, I see no reason why {{cite web}} cannot underlie {{cite FTP}}.

{{cite FTP}} must not provide free advertising for some company's product in en.wiki articles as its current incarnation will do.

—Trappist the monk (talk) 15:38, 4 May 2025 (UTC)

I linked instead to Comparison of FTP client software, no strong preference on link target. Regarding |archive-url=, I didn't realize until this discussion that Archive.org saves FTP links. Archive.today just gives an "Invalid URL: ftp://..." message. An archive link like, https://web.archive.org/web/20220121090419/ftp://naif.jpl.nasa.gov/pub/naif/generic_kernels/spk/planets/a_old_versions/de418_announcement.pdf, is probably accessible to a much broader range of users. If the plan is to have a bot convert these to an FTP template, could the bot also check for archives? No strong on opinions on other notes above, Rjjⁱⁱⁱ (talk) 16:00, 4 May 2025 (UTC)

Archive.org saves FTP links when they have an HTTPS interface/gateway. It's saving the https version of the link but displaying the URI as if it were ftp:// .. if the URI has no interface then archive providers can't save it. Ideally any ftp:// that has a gateway, we modify the URI to https:// but no guarantee Archive.org saved the URI as https:// they may have it saved the URI ftp:// even though it's actually through a https:// interface. Confusing. The purpose/need for Cite FTP template is to support ftp:// links that do not have an interface gateway, that are pure 100% FTP protocol and nothing else, no hybrid link. Those hybrids, we simply convert the URI to https:// and the FTP problem disappears, they can continue to use cite web. -- GreenC 18:21, 4 May 2025 (UTC)

What template is link a parameter of? Jc3s5h (talk) 16:10, 4 May 2025 (UTC)

The current incarnation of {{cite FTP}}; see the above examples, in particular: No link.

—Trappist the monk (talk) 16:17, 4 May 2025 (UTC)

Is this a wrapper for {{Citation}} or something else, such as {{Cite book}}? If the former, sometimes Citation doesn't guess correctly whether the publication being cited is a book, journal, etc. Is there a way to specify what kind of publication is being cited? Jc3s5h (talk) 15:09, 4 May 2025 (UTC)

In its current form, wrapping {{citation}}, {{cite FTP}} will require some sort of parameter (|server=?) to feed |work= so that {{citation}} renders |title= upright and quoted. If that parameter is not present, |title= will be rendered in italics.

—Trappist the monk (talk) 15:38, 4 May 2025 (UTC)

Trappist, would you be willing to modify or make a template that conforms to what you envision? As noted above, this template is primarily for FTP links that are pure 100% FTP protocol, that do not redirect to a HTTPS server via an interface/gateway. These types of "pure" FTP links will never work in a browser, without special modifications and software installed. For the vast majority of users they will get page not found. Of course this won't stop editors from adding interface FTP links that do work in browsers, thus the need for turning on or off the hyperlink, but in those cases I can't think of any reason why they would use Cite FTP and not Cite web, simply change the URI from ftp:// to https:// -- GreenC 18:29, 4 May 2025 (UTC)

See {{cite FTP/sandbox}}. Using the same example as above:

• {{cite FTP/sandbox |author=W. M. Folkner |author2=E. M. Standish |author3=J. G. Williams |author4=D. H. Boggs |date=2 August 2007 |title=Planetary and lunar ephemeris DE418 |server=Jet Propulsion Laboratory |url=ftp://naif.jpl.nasa.gov/pub/naif/generic_kernels/spk/planets/a_old_versions/de418_announcement.pdf}}

  W. M. Folkner; E. M. Standish; J. G. Williams; D. H. Boggs (2 August 2007). "Planetary and lunar ephemeris DE418" (PDF). Jet Propulsion Laboratory (FTP). (To view documents see Help:FTP.)

The obvious difference here is that |title= is linked and the template uses |server= instead of |work=. Another difference:

• {{cite FTP/sandbox |author=W. M. Folkner |author2=E. M. Standish |author3=J. G. Williams |author4=D. H. Boggs |date=2 August 2007 |title=Planetary and lunar ephemeris DE418 |server=Jet Propulsion Laboratory |url=http://naif.jpl.nasa.gov/pub/naif/generic_kernels/spk/planets/a_old_versions/de418_announcement.pdf}}

  W. M. Folkner; E. M. Standish; J. G. Williams; D. H. Boggs (2 August 2007). "Planetary and lunar ephemeris DE418" (PDF). Jet Propulsion Laboratory (FTP). {{Cite FTP}} must use ftp:// scheme (To view documents see Help:FTP.)

The template emits an error message when |url= holds sommat other than an ftp url; category applies only to mainspace. |server= isn't currently required but probably ought to be. There is no preview message.

Earlier in this discussion it was suggested that a help page should be linked from this template. At this writing, both the live and sandbox versions of the template link to Comparison of FTP client software which to my mind does not answer. The help text can be on a separate page in the Help namespace or can be included in the template documentation. I guess the latter would be my preference. I leave it to others to write that help text.

—Trappist the monk (talk) 22:47, 4 May 2025 (UTC)

User:Trappist the monk: This is fine, thank you. I agree the help page should be in Help: namespace, so editors can create custom instructions, and link to relevant pages internal and external. As noted by User:Jc3s5h below, the example FTP link is actually a dead FTP site. How do we handle those? Adding a {{dead link}} won't do anything with bots, nor will adding an |archive-url=, because the web archive providers (Wayback Machine, Archive.today etc) only work with web archives, not FTP. Perhaps the Help: page can have a section on this (searching for alternatives), though I'd still like some way to indicate the FTP site is inoperable. Any thoughts? Possibly a new version of {{dead link}} such as {{dead FTP}} which can be tracked separately. -- GreenC 18:39, 7 May 2025 (UTC)

I would much rather have a help page about ftp access rather than an article about ftp clients, because we can give practical advice without finding "reliable sources". Many users will be using Windows 11, and most of the information about ftp in Windows 11 is in question-and-answer forums rather than Microsoft documentation, or other reliable sources that could be cited in an article. My recommendation for the Windows 11 user who doesn't want to bother installing a special client is use Windows Explorer. For simple cases, just put the URL in the address bar of Windows Explorer. Jc3s5h (talk) 20:39, 7 May 2025 (UTC)

I have tweaked the ~/sandbox version so that it ignores |archive-date=, |archivedate=, |archive-url=, |archiveurl= and |url-status=. These parameters can be set so when a dead ftp is encountered, editors can set |url-status=dead ({{cite FTP/sandbox}} is a wrapper template so the cs1|2 rules can be bent). There is no visible indication but perhaps better than an unattached {{dead link}} template and perhaps bots can use it. Also changed ~/sandbox to link to Help:FTP:

• {{cite FTP/sandbox |author=W. M. Folkner |author2=E. M. Standish |author3=J. G. Williams |author4=D. H. Boggs |date=2 August 2007 |title=Planetary and lunar ephemeris DE418 |server=Jet Propulsion Laboratory |url=ftp://naif.jpl.nasa.gov/pub/naif/generic_kernels/spk/planets/a_old_versions/de418_announcement.pdf |url-status=dead}}

  W. M. Folkner; E. M. Standish; J. G. Williams; D. H. Boggs (2 August 2007). "Planetary and lunar ephemeris DE418" (PDF). Jet Propulsion Laboratory (FTP). (To view documents see Help:FTP.)

—Trappist the monk (talk) 15:37, 8 May 2025 (UTC)

Great. If url-status is set, is it possible to produce a visual notice like except "dead FTP link"? -- GreenC 16:22, 8 May 2025 (UTC)

Yeah, see the example above. Can be linked; if so, where? Mayhaps we should change the parameter to |ftp-status=dead?

—Trappist the monk (talk) 18:18, 8 May 2025 (UTC)

Trappist the monk that's perfect. I guess leave unlinked until something better comes along. According to URL anything with protocol://example.com is a URL/URI while something like "example.com" is not a URL but would be a URI. If we were concerned about using "url-status" there would also be concern with |url= .. however technically URL is correct. Suggest leave as url. -- GreenC 14:24, 11 May 2025 (UTC)

ok. I've added categorization for dead ftp links and tweaked the category names to: Category:Articles using cite FTP with improper URL and Category:Articles with dead FTP links. Categories created. Live template updated. Documentation tweaked. I'll write an AWB script today that will convert {{cite web}} with |url=ftp://... to {{cite FTP}}.

—Trappist the monk (talk) 16:01, 11 May 2025 (UTC)

OK. Conversion is more complex. Need to check each ftp:// and see if it has an HTTPS gateway available and if so convert the url to https:// which is preferable to retain the cite web. Make sure the https:// link works. If no gateway, then covert to {{cite FTP}} and run a FTP client (ncftpget) to verify the link is operational, updating ftp-status accordingly. Feel free to do this if you want, I was going to repurpose some code I wrote in 2021. -- GreenC 20:02, 11 May 2025 (UTC)

Oh, ok, you do that. I was just going to do a simple convert. Your way would be much better. Don't forget to change |publisher=, |website=, |work= to |server=.

—Trappist the monk (talk) 21:33, 11 May 2025 (UTC)

Does the ftp service on naif.jpl.nasa.gov actually work? I cannot connect to it. Is this example from a real Wikipedia article? Jc3s5h (talk) 23:10, 4 May 2025 (UTC)

Here is a working example, in the sense that the FTP server exists, and the file exists:

• {{cite FTP/sandbox |last1=Standish |first1=Myles |date=October 31, 1997 |title=CDroms |server=Jet Propulsion Laboratory |url=ftp://ssd.jpl.nasa.gov/pub/eph/planets/}}
• Standish, Myles (October 31, 1997). "CDroms". Jet Propulsion Laboratory (FTP). (To view documents see Help:FTP.)

Jc3s5h (talk) 23:50, 4 May 2025 (UTC)

I have edited Help:FTP so the example works. Jc3s5h (talk) 14:56, 11 May 2025 (UTC)

Why a cite ftp template, and not just url protocol recognition? Headbomb {t · c · p · b} 18:46, 11 May 2025 (UTC)

What do you mean by url protocol recognition? {{cite web}} recognizes that |url= has a url with an ftp:// scheme? Then what?

—Trappist the monk (talk) 18:55, 11 May 2025 (UTC)

that any of {{cite xxx}} recognizes the FTP scheme. If there's a journal article hosted on an FTP, the right template to use would still be {{cite journal|url=ftp://...}}. Same for a book chapter hosted on an ftp, {{cite book|chapter-url=ftp://...}}, and so on. Headbomb {t · c · p · b} 19:39, 11 May 2025 (UTC)

That was brought up further up, my response began: "Possibly. It could be messy.." In brief this is cleaner and easier to manage. I don't think the quantity and quality of links, and future of FTP, warrants integration. Looking at the links, it's over 50% NOAA. Which is currently being massively defunded. Throughout the federal government there are big changes afoot with computer systems, headed by DOGE ie. Musk's Silicon Valley wrecking crew, for technical reasons alone it's reasonable to assume they are going to deprecate FTP in line with the browser companies. There will always be some small number of holdouts. -- GreenC 20:21, 11 May 2025 (UTC)

Is there a reason Template:CS1 language sources uses Module:Cs1 documentation support/sandbox?

Looking into supporting {{CS1 language sources}} categories for Wikipedia:Bots/Requests for approval/AnomieBOT 84, I found that {{CS1 language sources/core}} invokes a function from Module:Cs1 documentation support/sandbox rather than Module:Cs1 documentation support. Is there a reason the sandbox hasn't been synced to the main module and the template updated to use it instead of the sandbox? If not, I'll probably go ahead and do that. Anomie⚔ 02:29, 3 May 2025 (UTC)

That comes from this edit 17 November 2024. Editor Gonnym?

—Trappist the monk (talk) 02:45, 3 May 2025 (UTC)

No idea why I left it like this. If there are no issues you see with it, I can sync with the live verison. Gonnym (talk) 18:20, 4 May 2025 (UTC)

Citation templates script-title

For the various {{citation}}/cite templates, could we make it so that script-title is an adequate replacement for "title"? When referencing a non-English source, to my understanding it's more appropriate to use script-title and not title, but when doing so I get a warning message in VisualEditor that tells me that title is required. seefooddiet (talk) 14:42, 4 May 2025 (UTC)

This is not a failing of cs1|2. But, it is yet another failing of visual editor and templatedata. All cs1|2 templates accept |script-title= regardless of the presence or absence of |title=. If you believe differently, please give a real example of such failing.

—Trappist the monk (talk) 18:30, 4 May 2025 (UTC)

I had suspected it was a problem of visual editor and template data. Is it possible to fix/work around? To my knowledge it isn't. seefooddiet (talk) 19:45, 4 May 2025 (UTC)

In these cases, besides putting the original title in |script-title=, it is often useful to put a romanization of the title in |title=. Kanguole 19:51, 4 May 2025 (UTC)

Imo romanizations of titles are often excessively long and useful to few (difficult to read). Instead, I use |trans-title=. But if you use that param instead of title it still triggers the warning in VisualEditor. seefooddiet (talk) 19:57, 4 May 2025 (UTC)

I typically use all three, but it is true that Chinese and Japanese romanizations are more useful than Korean ones. But the reference list is out of the flow of the article text, so space is not at a premium. Kanguole 20:06, 4 May 2025 (UTC)

I'll still respectfully withhold from using it though; it's an extremely rare practice for Korea-related articles on enwiki. I've basically never seen it done. seefooddiet (talk) 20:10, 4 May 2025 (UTC)

Trans-series?

It has already been suggested in an edit request, but would it be useful to add a translated version of the paramater series; trans-series? Similar to title and chapter. I have a series of books ("Parlementaire geschiedenis van Nederland na 1945"), and it would be helpful for English readers to know what this series is about. Dajasj (talk) 08:50, 6 May 2025 (UTC)

Parameters for English translations of foreign sources?

When citing an English translation, how do you indicate the original language, title and chapter title? The documentation for {{citation}} seems to suggest that the obvious parameters are only for use when citing in the original language. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 12:31, 7 May 2025 (UTC)

I would refer to a separate {{citation}} for the original work after the translation. Kanguole 21:00, 7 May 2025 (UTC)

What if all I have is the translation? -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 14:22, 8 May 2025 (UTC)

For {{Cite book}}, if you want to indicate the original title there is |trans-title= and |trans-chapter=, or the original publication date |orig-date= or year |orig-year=. The translator is |translator-last= and |translator-first=. Usually these (trans-* and orig-*) are sufficient. There is no |orig-lang= option, don't know why. -- GreenC 00:21, 8 May 2025 (UTC)

That's what I would assume given only the names of the parameters, but the documentation says things like trans-title: If the cited source is in a foreign language,. In the case I'm concerned with, the cited source is the translation. Thus, for this citation, |title=פִּרְקֵי אָבוֹת and |trans-title=Sayings of the Jewish Fathers (Pirqe Aboth) would imply that I was citing the original rather than the translation, as would |script-title=פִּרְקֵי אָבוֹת and |trans-title=Sayings of the Jewish Fathers (Pirqe Aboth). There's also the problem that the English title is not actually a translation of the original Hebrew. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 14:22, 8 May 2025 (UTC)

The documentation for |trans-title= says (in part):

trans-title: English translation of the title if the source cited is in a foreign language.

If you are citing the source at sacred-texts.com, do not include |script-title=he:פִּרְקֵי אָבוֹת because that text does not appear to be present in that source. Do not shoehorn multiple sources into a single cs1|2 template; the templates are designed (if we can use that term) to support only one source at a time.

—Trappist the monk (talk) 15:12, 8 May 2025 (UTC)

References

Template:Cite book#csdoc_trans_title

"Module:Citation" listed at Redirects for discussion

The redirect Module:Citation has been listed at redirects for discussion to determine whether its use and function meets the redirect guidelines. Readers of this page are welcome to comment on this redirect at Wikipedia:Redirects for discussion/Log/2025 April 14#Module:Citation until a consensus is reached. Rusalkii (talk) 22:17, 7 May 2025 (UTC)

ignored parameters not ignored

I recently found this template:

{{cite SSRN |first1=Amanda |last1=Glazer |first2=Jacob |last2=Spertus |orig-date=8 March 2020 |date=2020-02-10 |df=dmy-all |title=Start spreading the news: New York's post-election audit has major flaws |ssrn=3536011}}

Glazer, Amanda; Spertus, Jacob (10 February 2020) [8 March 2020]. "Start spreading the news: New York's post-election audit has major flaws". SSRN 3536011. {{cite SSRN}}: Unknown parameter |orig-date= ignored (help)

The error message says that |orig-date= should be ignored yet there it is in the rendering.

Module:Citation gets parameters from the parent frame's args table. Though this table cannot be modified, the code attempted to modify it by setting the value assigned args['orig-date'] = ''. Fixed in the sandbox:

{{cite SSRN/new |first1=Amanda |last1=Glazer |first2=Jacob |last2=Spertus |orig-date=8 March 2020 |date=2020-02-10 |df=dmy-all |title=Start spreading the news: New York's post-election audit has major flaws |ssrn=3536011}}

Glazer, Amanda; Spertus, Jacob (10 February 2020). "Start spreading the news: New York's post-election audit has major flaws". SSRN 3536011. {{cite SSRN}}: Unknown parameter |orig-date= ignored (help)

—Trappist the monk (talk) 15:01, 14 May 2025 (UTC)

3-digit year makes error

• {{cite book|title=title|date=Jan 500}}
• title. Jan 500. {{cite book}}: Check date values in: |date= (help)
• {{cite book|title=title|date=Jan 0500}}
• title. Jan 0500.

The former will result in an error, but the latter will not. This doesn't seem to comply with MOS:BADDATE, but is this intentional? What is the meaning of difference between (%d%d%d%d) and (%d%d%d%d?) in patterns_t of /Date validation? --FlatLanguage (talk) 02:38, 15 May 2025 (UTC)

It is intentional. Please cite the version that you are actually looking at. The assumption in the code is that you are not citing an actual physical book that was created in AD 500. You can use |orig-year=500 for the original publication date. – Jonesey95 (talk) 15:09, 17 May 2025 (UTC)

I thought that I had replied to this... Alas, it appears not.

Is it even possible to know the publication date to month precision, for such an old publication? Regardless, the leading zero version (|date=Jan 0500) above should also emit an error message. Further, is there any reason why a leading zero should be allowed in the year portion of a publication date?

—Trappist the monk (talk) 15:20, 17 May 2025 (UTC)

Since I'm only localizing it for jawiki, the specifications can be decided by the people at enwiki. Now Mdy, dMy, y-y and y allow 3-digits year, and the others (ymd, Md-dy, d-dMy, dM-dMy, Md-Mdy, dMy-dMy, Mdy-Mdy, My-My, M-My, My, Sy4-y2, Sy-y, y4-y2) don't. All allow leading zero version. --FlatLanguage (talk) 15:44, 17 May 2025 (UTC)

The conformance (and lack thereof) of the citation templates to WP:MOSNUM is documented at Help:Citation Style 1#Date format compliance with Wikipedia's Manual of Style. I contend we should follow the wisdom of Fred Brooks who in his book The Mythical Man-Month that the designers of OS/360 learned the hard way, it is essential that the behavior of software match its documentation.

A fault with the documentation is that it fails to explain what an exception to compliance is. Is it

• The template renders a falsehood?
• The template ignores the parameter?
• The template renders a true result together with an error message?

So I added the following sentence to "Citation Style 1"

When the template displays an error message (to a reader who has not chosen enhanced errors or warnings) for a value that complies with Wikipedia:Manual of Style/Dates and numbers § Dates, months, and years the template is considered non-compliant for that value.

Of course this will either require changing the table to show values that produce error messages, or changing my sentence to indicate that a correct rendered value along with an error message is compliant. Jc3s5h (talk) 16:34, 17 May 2025 (UTC)

@Jonesey95: The code should follow the documentation, or at least provide a truthful escape hatch. The documentation says three digit years are allowed. Unfortunately, the documentation isn't limited to citations, and of course there is a need to use three digit years in the running text of articles. I suggest some sort of escape hatch be provided, such as |date=((May 950)). It is wrong to use |orig-year= if one actually read the original, or a scan of the original available from a reliable source, such as Oxford University. Jc3s5h (talk) 16:44, 17 May 2025 (UTC)

Trappist the Monk: WP:MOSNUM specifically includes June 0622 in the unacceptable date formats table.

Question for Trappist or others: Where can we add these or similar examples to the test cases? Jc3s5h (talk) 16:51, 17 May 2025 (UTC)

Historically, cs1|2 documentation lags the code. The documentation is not protected. If you see where and how the documentation can be improved, please do so. A good place to start might be to confirm that Help:Citation Style 1 § Date format compliance with Wikipedia's Manual of Style correctly lists compliance with and exceptions to MOS:DATE. For example, the Acceptable date formats table identifies '2 Sep' and 'Sep 2' (dates without years) as acceptable but cs1|2 will reject those sorts of dates though the Date format compliance with Wikipedia's Manual of Style does not say so.

Date-handling testcases are defined at Module:Citation/CS1/testcases/dates with the results rendered at Module talk:Citation/CS1/testcases/dates.

Thanks for the June 0622 pointer; I'll fix the module to reject zero-padded years.

Still unanswered is my question about three-digit-year publication dates with month or month-day precision. It is pointed out above that we are somewhat inconsistent in our support of three-digit years. Should we really be supporting such dates?

—Trappist the monk (talk) 18:49, 17 May 2025 (UTC)

Advanced searches of Wikipedia are not my strong suit; when I've tried to use Advanced search in Special:Search the finds and misses often mystified me. Can anyone suggest how to search for pages that cite works from the 1st millennium? Jc3s5h (talk) 19:05, 17 May 2025 (UTC)

Here are three searches:

|year=### – times out; finds ~100 articles

|date=### – times out; finds ~70 articles

|date=mmmmm ### – times out; finds no articles

—Trappist the monk (talk) 19:43, 17 May 2025 (UTC)

Sandbox now rejects zero-padded year values:

{{cite book/new|title=title|date=0500}} → title. 0500. {{cite book}}: Check date values in: |date= (help)

{{cite book/new|title=title|date=Jan 0500}} → title. Jan 0500. {{cite book}}: Check date values in: |date= (help)

—Trappist the monk (talk) 19:43, 17 May 2025 (UTC)

In check_date(), only year2 is checked by is_valid_year() for My-My and y-y, so

Cite book comparison

Wikitext	{{cite book|date=Jan 0999 – Jan 1000|title=title}}
Live	title. Jan 0999 – Jan 1000.
Sandbox	title. Jan 0999 – Jan 1000.

Cite book comparison

Wikitext	{{cite book|date=0999–1000|title=title}}
Live	title. 0999–1000.
Sandbox	title. 0999–1000.

I think good way is to use [1-9]%d%d%d? in patterns_t except ymd.

--FlatLanguage (talk) 02:40, 18 May 2025 (UTC)

FlatLanguage, is it your intent to not support three digit years, or would those be handled somewhere else? Jc3s5h (talk) 15:49, 18 May 2025 (UTC)

@Jc3s5h: I don't understand the meaning of your question. Trappist the monk fixed sandbox of the module to make 0YYY error, but I pointed out it overlook 0YYY–YYYY.

Now the module use pattern %d%d%d%d or %d%d%d%d? and what I suggested was [1-9]%d%d%d?. %d means digit, [1-9] means 1 to 9 digit, and '?' means that the previous character is optional. So [1-9]%d%d%d? means 3 or 4 digits led by a non-zero number. FlatLanguage (talk) 01:16, 19 May 2025 (UTC)

Thanks, I was expecting the first digit to be optional, but it's the last digit that's optional, which gives the same result. Jc3s5h (talk) 14:58, 19 May 2025 (UTC)

date: 2000-10-01, year: 2010 doesn't make error

• {{cite book|title=title|date=1 Oct 2000|year=2010}}
• title. 1 Oct 2000. {{cite book}}: Check date values in: |year= / |date= mismatch (help)
• {{cite book|title=title|date=2000-10-01|year=2010}}
• title. 2000-10-01.{{cite book}}: CS1 maint: date and year (link)

A part of year_date_check() in /Date_validation is folowing:

	if date_string:match ('%d%d%d%d%-%d%d%-%d%d') and year_string:match ('%d%d%d%d%a') then	--special case where both date and year are required YYYY-MM-DD and YYYYx
		date1 = date_string:match ('(%d%d%d%d)');
		year = year_string:match ('(%d%d%d%d)');
		if year ~= date1 then
			result = 0;															-- years don't match
		else
			result = 2;															-- years match; but because disambiguated, don't add to maint cat
		end

It should be fixed like following:

	if date_string:match ('%d%d%d%d%-%d%d%-%d%d') then	--special case where both date and year are required YYYY-MM-DD and YYYYx
		date1 = date_string:match ('(%d%d%d%d)');
		year = year_string:match ('(%d%d%d%d)');
		if year ~= date1 then
			result = 0;															-- years don't match
		elseif year_string:match ('%d%d%d%d%a')
			result = 2;															-- years match; but because disambiguated, don't add to maint cat
		end

Or, exclude YYYY-MM-DD at later branch of YYYY-YY. --FlatLanguage (talk) 02:56, 15 May 2025 (UTC)

Or, much easier, flag as an error a citation that attempts to use both |date= and |year=. 𝕁𝕄𝔽 (talk) 09:32, 15 May 2025 (UTC)

Isn't that Category:CS1 maint: date and year? CMD (talk) 12:37, 15 May 2025 (UTC)

Yes, the problem I pointed out is that the latter should be err_bad_date, not maint_date_and_year. Now 2000-10-01 is parsed as 2000–2010 in year_date_check(). --FlatLanguage (talk) 13:13, 15 May 2025 (UTC)

As I recall, when citing two works by the same author in the same year, and at least one of the works had a date more precise than the year, it was necessary to cite the date as, for example, |date=January 18, 2015 |year=2015b. Now, the "b" could be added to the date parameter, |date=January 18, 2015b, but that wasn't always the case. If the publication dates were all-numeric format, one solution would be to write {para|date|2015-01-18}} |year=2015b. It wouldn't work to use |date=2015b-01-18. (My solution would be to change every all-numeric date in all the citations to a format with a spelled-out month since ISO 8601 is not really suitable for Wikipedia articles.) Jc3s5h (talk) 14:41, 15 May 2025 (UTC)

ISO dates are fine for the vast majority of articles. Headbomb {t · c · p · b} 19:31, 17 May 2025 (UTC)

The YYYY–YY test in the live module isn't correctly bounded. Because of that, the the test extracts the YYYY-MM portion of |date=2000-10-01 believing it to be an abbreviated year range YYYY–YY (2000–10). The test constructs a 4-digit year from the century digits from 2000 (20) and the MM digits (10) of what it thought was the abbreviated year to make YYMM → 2010. Because 2010 from 2010-10-01 matches the constructed value (2010), the test emits the maintenance message instead of the error message. Fixed in the sandbox by properly bounding the YYYY–YY pattern (line 823 (permalink)):

Cite book comparison

Wikitext	{{cite book|date=2000-10-01|title=title|year=2000}}
Live	title. 2000-10-01.{{cite book}}: CS1 maint: date and year (link)
Sandbox	title. 2000-10-01.{{cite book}}: CS1 maint: date and year (link)

Cite book comparison

Wikitext	{{cite book|date=2000-10-01|title=title|year=2010}}
Live	title. 2000-10-01.{{cite book}}: CS1 maint: date and year (link)
Sandbox	title. 2000-10-01. {{cite book}}: Check date values in: |year= / |date= mismatch (help)

—Trappist the monk (talk) 22:15, 17 May 2025 (UTC)

Now Sy4-y2 recognized incorrectly:

Cite book comparison

Wikitext	{{cite book|date=Winter 2000–01|title=title|year=2001}}
Live	title. Winter 2000–01.{{cite book}}: CS1 maint: date and year (link)
Sandbox	title. Winter 2000–01. {{cite book}}: Check date values in: |year= / |date= mismatch (help)

I think a good way is to handle all of the YYYY-MM-DD in the first branch, as I indicated at the beginning. FlatLanguage (talk) 02:10, 18 May 2025 (UTC)

trans-title is not included in the link

Template:Cite web#Title says: "trans-title is included in the link". I interpret this to mean that the value of the parameter is clickable but it's not as seen at Template:Cite web#Foreign language and translated title. Is the documentation wrong or did I misunderstand something? I don't think it should be clickable. I'm just commenting on the documentation. PrimeHunter (talk) 14:22, 16 May 2025 (UTC)

The decision to exclude |title-link= from the |title= external link was taken at Help talk:Citation Style 1/Archive 30 § Stop linking trans-title. The documentation was never updated. Done now.

—Trappist the monk (talk) 15:56, 16 May 2025 (UTC)

Bloat

reference info for Hip-hop

unnamed refs	242
named refs	74
self closed	32
	1
cs1 refs	243
cs1 templates	214
rp templates	20
cleanup templates	12
dead link templates	2
webarchive templates	16
use xxx dates	mdy
cs1|2 mdy dates	124
cs1|2 ymd dates	13
cs1|2 mdy access dates	149
cs1|2 ymd access dates	22
cs1|2 mdy archive dates	92
cs1|2 last/first	108
cs1|2 author	17
List of cs1 templates cite book (20) cite encyclopedia (1) cite journal (11) cite magazine (32) cite news (38) cite web (112)
explanations

Citation templates are great tools for editors, but they do create massive bloat that accretes to real harm. In an article like hip-hop, converting just one Cite Journal reference to plain text saved 360 bytes. The entire article is 226,834 bytes. An enormous amount of its current bloat is a reliance on templates for citations where they are not necessary. If there were a way to get templates to generate more efficient citations, that would be a blessing.Trumpetrep (talk) 21:04, 18 May 2025 (UTC)

If you are referring to this edit, please read WP:CITEVAR and MOS:CURLY to see why that edit was probably not in conformance with Wikipedia's guidelines. The majority of that article's references appear to use CS1 templates. Citation templates have both costs and benefits. It appears that you are focused only on the costs. – Jonesey95 (talk) 21:29, 20 May 2025 (UTC)

If you're concerned about bloat, you can remove some of the automatically prefilled parameters which do not add any value for a {{cite journal}}, starting from the hardcoded URL. Instead of

{{Cite journal|last=Leach|first=Andrew|year=2008|title="One Day It'll All Make Sense": Hip-Hop and Rap Resources for Music Librarians|url=https://www.jstor.org/stable/30163606|journal=Notes|volume=65|issue=1|pages=9–37|doi=10.1353/not.0.0039|jstor=30163606|s2cid=144572911|issn=0027-4380|access-date=December 5, 2020|archive-date=January 28, 2021|archive-url=https://web.archive.org/web/20210128051605/https://www.jstor.org/stable/30163606|url-status=live}}

it's enough to have

{{Cite journal|last=Leach|first=Andrew|year=2008|title="One Day It'll All Make Sense": Hip-Hop and Rap Resources for Music Librarians|journal=Notes|volume=65|issue=1|pages=9–37|doi=10.1353/not.0.0039|jstor=30163606}}

Nemo 13:02, 22 May 2025 (UTC)

Finding invisible characters

I'm looking to address a few entries in Category:CS1 errors: invisible characters but am having a few issues. Take Joel Olson for instance. (1) The page is listed in the above category page but the category is not showing on Olson's page for me, even though I have other hidden categories visible. (2) When I put the page into a hidden character viewer, I'm not seeing any characters afoul. Why is it still listed in that error category? czar 02:35, 21 May 2025 (UTC)

Joel Olson is not in that category... Headbomb {t · c · p · b} 03:08, 21 May 2025 (UTC)

I see it in the category, under O for Olson. CMD (talk) 07:17, 21 May 2025 (UTC)

Um right, so I did see it when CZAR first posted this. Checking now, it is no longer there. CMD (talk) 07:18, 21 May 2025 (UTC)

Sometimes the categories don't update immediately, purging the article or doing a dummy edit will usually force the update. -- LCU ActivelyDisinterested «@» °∆t° 09:09, 21 May 2025 (UTC)

I'd buy this if the article had been edited since last Christmas. Maybe a template or something the article used was edited. CMD (talk) 09:12, 21 May 2025 (UTC)

I've gone and purged the category (and took the opportunity to fix some hard coded nonbreaking spaces), members went down from ~2200 to ~250. I'm now running Citation bot on the category to handle a few more cases. Headbomb {t · c · p · b} 12:22, 21 May 2025 (UTC)

Thanks, all! czar 13:19, 21 May 2025 (UTC)

Now down to ~80some citations. Much easier to tackle the rest by hand (though it won't be me). Headbomb {t · c · p · b} 16:22, 21 May 2025 (UTC)

These type of problems - pages showing (or not showing) categories at the bottom, and category pages listing (or not listing) pages, is down to the link tables. A WP:PURGE will not update those, but a WP:NULLEDIT normally will. This needs to be performed on the individual pages, not on the category page. --Redrose64 🌹 (talk) 17:47, 21 May 2025 (UTC)

That's what I did. I suppose I could have been a bit clearer. Headbomb {t · c · p · b} 17:55, 21 May 2025 (UTC)

I don't see why we have this category at all. Surely if the templates are capable of observing the existence of invisible characters they are capable of stripping them out without causing errors. —David Eppstein (talk) 17:32, 21 May 2025 (UTC)

It prevents copy-pasting and search issues, essentially, though the category trackings other issues as well, like stray line breaks and stuff. Headbomb {t · c · p · b} 17:59, 21 May 2025 (UTC)

It's weird to display like 'In (someone). (book title)'

{{cite book |last=Bloggs |first=Fred |date=January 1, 2001 |editor-last=Doe |editor-first=John |title=Big Compilation Book with Many Chapters and Distinct Chapter Authors |publisher=Book Publishers |pages=100–110 |chapter=Chapter 2: The History of the Bloggs Family |isbn=}}

1. Bloggs, Fred (January 1, 2001). "Chapter 2: The History of the Bloggs Family". In Doe, John (ed.). Big Compilation Book with Many Chapters and Distinct Chapter Authors. Book Publishers. pp. 100–110.

Is it just me who think to be weird coming 'In' in front of editor name, when only coming up with the chapter parameter? Watsondoe (talk) 09:53, 21 May 2025 (UTC)

Especially in cite book Watsondoe (talk) 09:57, 21 May 2025 (UTC)

It looks like you are calling the APA style weird. Chicago does it differently, but a style has to settle on something consistent. – Jonesey95 (talk) 13:58, 21 May 2025 (UTC)

Yes, it's weird, but off the top of my head I can't think of an alternative that isn't also klunky in one way or another. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 15:49, 21 May 2025 (UTC)