Help talk:Citation Style 1/Archive 96

Archive 90

←

Archive 94

Archive 95

Archive 96

{Cite journal} pages styling

There is no 'p.', 'pg.', 'p ' or something of the sort indicating the numbers supplied in the page= or pages= parameters are pages for {cite journal}. Please change the displayed text from '<supplied-numbers>' to 'p. <supplied numbers>'.

Observe the example below, where I have only included page=400-500. It only displays '400-500', but it should display 'p. 400-500'

: 400-500. {{cite journal}}: Cite journal requires |journal= (help); Missing or empty |title= (help)

(Perhaps I am wrong about this, and there is a special academic standard for citing journal page numbers I am unaware of?) Tule-hog (talk) 19:56, 4 August 2024 (UTC)

It seems fine to me as it is. What apart from a page span could be meant by an ndash-separated pair of increasing numbers following a colon following the volume and issue? Why not use {{cite periodical}} if you're unhappy with the display formatting of {{cite journal}}? Folly Mox (talk) 20:23, 4 August 2024 (UTC)

Firstly, note this also applies to page= which does not provide an n-dash. I thought the following was a little confusing along side a year:

. 2000: 1900. {{cite journal}}: Cite journal requires |journal= (help); Missing or empty |title= (help)

I thought it was incorrect, not bad. If others disagree, then let it stay! Tule-hog (talk) 20:49, 4 August 2024 (UTC)

Since we purport to be a publication for a generalist audience, I would think that we should format our citations for a generalist audience as well. Yes, that formatting is used in specialist publications, but we are not a specialist publication. For that reason, I continue to advocate that we should drop the overly truncated formatting for the volume/issue/page triplet on journals in favor of adding "vol.", "no." and "p." or "pp.". In short, it should be the way {{cite magazine}} does it without separating the vol./no. from the p. or pp. Imzadi 1979 → 20:57, 4 August 2024 (UTC)

I feel like the "vol." and "no." add undesirable cruft that is more difficult to identify on vgrep than the bolded volume number of {{cite journal}}. Just because we're not a specialist publication doesn't mean we shouldn't support multiple citation formats. Also, I like that the formatting between {{cite journal}} and {{cite magazine}} differs, which lets me see at a glance how reputable and peer-reviewed a source is / purports to be.

If there are confusion points in our citations to expunge for immediate understanding by any reader, I'd start with the lowest hanging fruit of Bibcodes and s2cids, which seldom if ever add any useful information even on clickthrough, and are entirely unfamiliar to almost every reader. Folly Mox (talk) 16:17, 5 August 2024 (UTC)

People get used to this formatting after seeing it a few times, and the concision is helpful. Making it more explicit helps someone the first time, but only at the expense of significantly reducing information density of reference sections, which isn't worth it IMO. –jacobolus (t) 02:36, 9 August 2024 (UTC)

 Not done for now: please establish a consensus for this alteration before using the {{Edit semi-protected}} template. – Jonesey95 (talk) 20:41, 4 August 2024 (UTC)

Is there a template to indicate the intent to use the {edit semi-protected} in future (i.e. 'request proposal', 'needs consensus', or something)? Tule-hog (talk) 20:52, 4 August 2024 (UTC)

No. You just discuss until there is consensus, and then you can remove |answered=yes from the template in this section to reactivate the request. – Jonesey95 (talk) 13:22, 5 August 2024 (UTC)

If an italicized title ends in a question mark, shouldn't the following punctuation be suppressed?

For example:

• Bryant, John; Sangwin, Chris (2008). How Round Is Your Circle?. Princeton University Press. ISBN 978-0-691-14992-9.

Or likewise with an exclamation mark:

• Eastman, P. D. (1961). Go, Dog. Go!. Random House. ISBN 0-394-80020-6.

It seems like the output should not put a period or comma immediately after another punctuation symbol.

Edit: the relevant manual of style section is MOS:CONSECUTIVE. –jacobolus (t) 06:37, 13 August 2024 (UTC)

Fully-protected edit request, August 17, 2024

You can see in the screenshots above that some characters (in this case, a lowercase I), when italicized, collide with the closing square brackets when using the |trans-title= parameter. This occurs on all skins. In this case I was using {{cite book}}, which italicizes the translated titles automatically, but this will also occur with other citation templates if the final character happens to be italicized. I have no idea where exactly this change needs to be made, but hopefully someone more well-versed in these modules will know better than me about that. Let me know if you have any questions! —TechnoSquirrel69 (sigh) 04:25, 17 August 2024 (UTC)

 Not done: it's not clear what changes you want to be made. Please mention the specific changes in a "change X to Y" format and provide a reliable source if appropriate. You didn't link to a page where this is happening, so I made my own. I get a slight white separation between the "i" and the right bracket. My default font is Lucida Grande, on Mac OS on a 2024-vintage laptop. I expect that everyone's formatting mileage will vary with font choices, screen resolution, and other computer-specific settings. Also, FWIW, your screen shots don't really show a connection or a gap, because of the blue highlighting. – Jonesey95 (talk) 05:12, 17 August 2024 (UTC)

@Jonesey95: Thanks, I'm still learning exactly what information is helpful in bug reports like this. I took those screenshots in Firefox on Vector 2022 while logged out, so all the formatting should be what the majority of readers will be seeing. I highlighted the I because it looked like the white dot clipping the top of the bracket more clearly illustrated the issue, but oh well. I'm seeing the same kerning issue in your sandbox, obviously, so perhaps we could do with a third opinion on whether this is worth addressing — I remember working with Trappist the monk on a similar issue a few months ago. I'm setting the request to open again. —TechnoSquirrel69 (sigh) 05:24, 17 August 2024 (UTC)

I won't edit war with you over the edit request template, but please read its instructions. There is no "complete and specific description of the request, so that an editor unfamiliar with the subject matter could complete the requested edit immediately" here. This is simply a discussion.

I am also using Firefox with the Vector 2022 skin. I don't think the problem here is with this particular template, or even with Wikipedia; it is with the way that an italic "i" is rendered when it is followed by a square bracket. On your computer, if you go to a text editor or other program where you can type, and set the font and font size to the same settings as Firefox, do you see the same problem? – Jonesey95 (talk) 13:26, 17 August 2024 (UTC)

This is fully dependent on client configuration (specifically, the kerning pairs in the chosen display typeface). A hair space output when transitioning from oblique text to upright closing bracket could accommodate problem typefaces, but might make display wonkier for typefaces where the interaction is already accounted for. Folly Mox (talk) 17:18, 17 August 2024 (UTC)

Not done, please made you request in a way that's actionable, e.g. "Change [this] to [that]". Until an exact change is requested, the edit request is premature. Headbomb {t · c · p · b} 13:50, 17 August 2024 (UTC)

Are Zotero Wikipedia Templates useable?

I just installed the "Wikipedia Templates" (version from 2021) from within Zotero, which are the only Wikipedia citation style that Zotero seems to have on offer. However, this always gives me a syntax that uses the "vauthors" parameter and always puts author names in double parentheses, e.g., vauthors=((Ellenberger, P.)). Is this correct? Am I doing something wrong, or is there an alternative Zotero citation style available somewhere that uses the standard last= and first= parameters instead? Thanks. Jens Lallensack (talk) 14:27, 23 August 2024 (UTC)

Abusing the accept-as-written markup is wrong. I seem to recall a conversation some place where the participants were trying to find a way to avoid multiple author names in |author= (or something like that). Instead of working out a proper solution (enumerated author-name parameters), they chose to abuse the accept-as-written markup as their solution.

If we are to believe this search, there are 500ish articles that have |vauthors= with multiple accept-as-written marked-up names. No doubt, there is a proportion of those articles that legitimately use the markup. Those that are not should be repaired.

I know nothing about Zotero or whatever it is that you installed. But, given that the Zotero thingy does not produce proper output, were I you, I would uninstall and complain to Zotero about the poor quality of their thingy. I would also not hold my breath waiting for a positive (any?) response.

—Trappist the monk (talk) 15:02, 23 August 2024 (UTC)

Thanks, that confirms my hunch that the Zotero templates are currently broken. I will try to fix it myself. Jens Lallensack (talk) 15:13, 23 August 2024 (UTC)

Template:Cite journal documentation states that (( )) should only be used for institutional authors. This may be a fine distinction Zotero is having trouble picking up. CMD (talk) 15:22, 23 August 2024 (UTC)

I just updated to the latest version, prompted by Jens Lallensack's post. I find the structure strange, even though I've been using it occasionally for years. In Edit->Settings, there is a Cite panel, which (by default) doesn't mention Wikipedia. But in the Edit->Settings->Export panel there is a Quick Copy feature, and you can select the format for items that are quick copied. One of the choices is Wikipedia Citation Templates. I've found this method to be reasonable, although some touch-ups are usually needed after I copy. Clarification: I did not take any steps to install any additional Wikipedia format. Jc3s5h (talk) 15:47, 23 August 2024 (UTC)

@Jc3s5h: I thought that Zotero does not ship any Wikipedia template by default? Are you sure that it's not the old Wikipedia template you installed years ago (which works fine, as I remember)? If you go "Edit" -> "Settings" -> "Cite", what does the "Updated" field say for your Wikipedia template? If it is the older version that does not use vauthors, could you maybe send it to me? (Just go "Settings" -> "Cite" -> Select "Wikipedia Templates" -> click on "Style editor", and copy the xml code). That would be much appreciated, thank you. Jens Lallensack (talk) 16:04, 23 August 2024 (UTC)

Never mind, I found it (as you wrote, the "Wikipedia Citation Templates", not the "Wikipedia Templates" from the style repository). Great, that works for me now. Thanks. --Jens Lallensack (talk) 16:12, 23 August 2024 (UTC)

Limit for s2cid needs increasing

Note that the value 270801523 is correct for s2cid and is larger than the currently configured limit of 270000000. —Jonathan Bowen (talk) 15:07, 23 August 2024 (UTC)

Notice: discussions on WMDE sub-reference project in progress at Meta

The WikiMedia sub-referencing project (parent project: Reusing references) is having multiple discussions about the development of a sub-referencing feature by WikiMedia Deutschland Engineering. Your feedback would be welcome at any of the discussons at m:Talk:WMDE Technical Wishes/Sub-referencing. Mathglot (talk) 21:00, 23 August 2024 (UTC)

Citoid adds PMC prefix

I sometimes make citations with "pmc" in the |pmc= parameter.[1] This is always the result of using Source Editor > Ref Toolbar > Cite > Templates > Cite journal > Autofill (Magnifying glass button). I believe the autofill is done by Citoid. Is it an error to include "pmc"? If so, how do we let the appropriate folks know? Rjjⁱⁱⁱ (talk) 20:47, 24 August 2024 (UTC)

I don't use ve so can't speak to users who do. When I look at the RefToolbar form for {{cite journal}}, PMC is hidden behind the Show/Hide extra fields button and does not have autofill (no quizzing glass). Filling with PMID (PMID 33061789 in this case) which does have a quizzing glass and which would be most likely autofillable identifier to add PMC, does not fill the PMC parameter. Filling from URL and DOI likewise do not fill PMC.

Are you sure that RefToolbar is how you created that {{cite journal}} template?

It is not an error to include the PMC prefix when filling |pmc= but that does cause cs1|2 to add the article to Category:CS1 maint: PMC format.

—Trappist the monk (talk) 21:42, 24 August 2024 (UTC)

Oh, dang, you're right, I must have done ProveIt on that one.

This doi "10.1098/rspa.2020.0148" in either the Visual Editor's automatic citation or ProveIt's automatic {{cite journal}} will add the prefix. Rjjⁱⁱⁱ (talk) 21:58, 24 August 2024 (UTC)

How do you use the autogenerated anchor, eg CITEREFBaldwinMarshall1999

In Toxic heavy metal the is an anchor CITEREFBaldwinMarshall1999, but I can't figure out how to use it in a ref tag. Johnjbarton (talk) 18:59, 25 August 2024 (UTC)

You can reference it with the {{sfn}} or {{harv}} families of templates. Kanguole 19:14, 25 August 2024 (UTC)

@Kanguole Thanks! But.. I'm still stuck. The sfn template needs names and dates. Where do I put "CITEREFBaldwinMarshall1999"? Thanks! Johnjbarton (talk) 19:22, 25 August 2024 (UTC)

The basic format for {{sfn}} templates for that would be {{sfn|Baldwin|Marshall|1999}}. -- LCU ActivelyDisinterested «@» °∆t° 19:31, 25 August 2024 (UTC)

The template automatically links to that format to CITEREFBaldwinMarshall1999. -- LCU ActivelyDisinterested «@» °∆t° 19:33, 25 August 2024 (UTC)

There are no {{sfn}} or {{harv}} templates in that article. Remember that WP:CITEVAR applies.

—Trappist the monk (talk) 19:37, 25 August 2024 (UTC)

That article is a mess. Keeping WP:CITEVAR in mind, to make a Baldwin Marshall 1999 short form reference, here are some options:

[[#CITEREFBaldwinMarshall1999|Baldwin Marshall 1999]]

[[#{{sfnref|Baldwin|Marshall|1999}}|Baldwin Marshall 1999]]

Both of these, to my mind, make life more difficult than it needs be. But, they work well with the cs1|2 templates. Alas, the article uses hand crafted fragment links so for strict compliance with WP:CITEVAR, this:

[[#Baldwin|Baldwin Marshall 1999]]

And you will get yet another short-form reference that does not link to its long-form companion – some do, but the majority do not so what is the purpose of all that fragment wikilinking?

Were I the chief, I would replace all of those fragment wikilinks with proper {{sfn}} templates. Alas, I'm not. To accomplish that you will need to achieve consensus at the article talk page

You are aware that there are two instances of Baldwin Marshall 1999 which appear to be the same source?

—Trappist the monk (talk) 19:37, 25 August 2024 (UTC)

Thanks, but to be honest I don't understand all these options and half of the words. Anyway someone else already changed over to sfn ;-) Johnjbarton (talk) 19:45, 25 August 2024 (UTC)

chapter-archive-url

Should there be a chapter-archive-url parameter for dead chapter-url? Tule-hog (talk) 21:33, 2 September 2024 (UTC)

Show a real-life example of where such a parameter is required. Note that when |archive-url= is provided in the presence of |chapter-url= with or without |url=, the module uses |archive-url= to link |chapter=:

{{cite book |chapter=Chapter |chapter-url=//example.com/chapter |title=Title |url=//example.com |archive-url=//archive.org |archive-date=2024-09-02}}

"Chapter". Title. Archived from the original on 2024-09-02.

If we are to support |chapter-archive-url= we must also support |chapter-archive-date=, and |chapter-url-status=.

Give an example to show how a template would render with |archive-url=, |archive-date=, |chapter-archive-url= and |chapter-archive-date=.

—Trappist the monk (talk) 21:57, 2 September 2024 (UTC)

Plus all the other *-url parameters. It's complex and messy. For those rare instances of ambiguity about which URL the archive-url is attached to, use {{webarchive}} with |format=addlarchive. -- GreenC 22:30, 2 September 2024 (UTC)

I'll use archive.org-esque convention of archive.org/<url> for clarity, omitting the date component. I'll keep close to your example.

{{cite book |chapter=Chapter |chapter-url=//example.com/chapter |chapter-archive-url=//archive.org/example.com/chapter |title=Title |url=//example.com |archive-url=//archive.org/example.com |archive-date=2024-09-02 |chapter-archive-date=2024-09-03 |chapter-url-status=dead}}

"Chapter". Title. Archived from the original on 2024-09-02, the section on 2024-09-03.

Example display is just a mockup, I'm not passionate about any particular method. Presumably chapter-url-status omitted by default would use chapter-archive-url, similar to url-status and archive-url. Tule-hog (talk) 22:32, 2 September 2024 (UTC)

CS1 translator feature request: addition of empty trans-title param

Can we add an empty trans-title param to the output of Module:CS1 translator, either always, or under control of a new, optional param?

I invariably try to follow up the module action by adding |trans-title=. Having that empty param added by the Module would simultaneously be a time-saver, as well as provide a quick check (via search-on-page + highlight) of how much work I still have left to do translating titles, and where they are located on the page. If this is accepted and implemented under parameteric control, please don't call the param 'trans', out of possible confusion with the whole purpose of the module; maybe something as simple as |_extra=y or |_expand=y or something.

The field even has utility empty: if I forget to come back and fill them in, some editor may come along and do it if they see the empty params, but may not look for the param in the doc if they don't know it exists.

Optionally, if additional languages such as Russian or Arabic or other non-Latin scripts are supported in the future, that same function or param could output the original title as |script-title= and add empty params |title= and |trans-title= ready for supplying romanization and translation. Mathglot (talk) 23:54, 4 September 2024 (UTC)

Just realized we have {{Cite journal/Arabic}}; does it do that? Mathglot (talk) 23:57, 4 September 2024 (UTC)

It does not.

—Trappist the monk (talk) 00:57, 5 September 2024 (UTC)

I'm not inclined to make this change primarily because some editor may come along and do it if they see the empty params whereupon, they grab the title from |title= and slap it into google translate, paste the translation into |trans-title= and claim: success! atta boy! job well done! Likely though, nothing will happen and the empty |trans-title= will remain empty; such is my experience.

If you want to find templates that Module:CS1 translator has translated, you can do an insource search for this:

<!-- auto-translated from <whatever language> by Module:CS1 translator -->

For example, this search for Arabic translations. Once you have visited a translated template and added/fixed as appropriate, delete the comment.

The purpose of the module is to (more-or-less accurately) do the grunt work of translating non-English template and parameter names to their English equivalents. Except for simple, mostly robust, month-name translations, that is all that the module does or should do.

—Trappist the monk (talk) 00:57, 5 September 2024 (UTC)

Another Generic Title

Hello, another Generic Title to be considered is "Unknown" or "unknown". Keith D (talk) 11:48, 5 September 2024 (UTC)

Feature request: author suffixes

Currently, an author name containing a comma-delimited suffix, e.g., George B. Thomas, Jr., generates the maintenance message CS1 maint: multiple names: authors list (link). This can be avoided by omitting the comma, e.g., |first1=George B. Jr or by wrapping in double parentheses, e.g., |first1=((George B., Jr)). A |suffixn= parameter would be a cleaner way to handle this.

Alternatively, the documentation should state that generational suffixes should not include the comma even if printed with a comma in the cited source. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 11:43, 9 September 2024 (UTC)

Isn't this already covered by MOS:JR? -- LCU ActivelyDisinterested «@» °∆t° 12:44, 9 September 2024 (UTC)

This is |last=Thomas+|first=George B. Jr to give Thomas, George B. Jr (2022). "Article of stuff". Journal of Things. 1 (2): 34–56. Headbomb {t · c · p · b} 12:51, 9 September 2024 (UTC)

Sorry, I incorrecly remembered Using Jr., Sr., or other such distinctions, including in the lead sentence of an article, is only for cases in which the name with the suffix is commonly used in reliable sources. as applying to the format of the suffix when it actually applies only to its existence. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 15:08, 9 September 2024 (UTC)

Errors at line 2083

Countless Lua errors at Gabrielle Baker: "Lua error in Module:Citation/CS1/Configuration at line 2083: attempt to index a boolean value." Even if it is an error in the article, it shouldn't give this error message I think. Fram (talk) 15:20, 10 September 2024 (UTC)

Are you sure? I don't see any error messages at Gabrielle Baker. It could be an artefact from the 2024-08-17 module suite update. During an update there is a brief period when new and old modules coexist; attempt to use the module suite in that brief period will show as lua script errors. Another possibility is that en.wiki could not interwiki to Commons to fetch identifier limits because of a fault at MediaWiki. The usual fix when there are many upon many of these sorts of error messages is a null edit or purge your cache.

—Trappist the monk (talk) 15:43, 10 September 2024 (UTC)

I checked through to see if anything was wrong, and made a couple of unrelated changes and the issue went away. It appears it needs a dummy edit for some reason. There are a few articles with the same error[2], one was Florida Man and a dummy edit cleared the issue. -- LCU ActivelyDisinterested «@» °∆t° 15:45, 10 September 2024 (UTC)

All but a few of the search results where already cleared, I've purged the others. -- LCU ActivelyDisinterested «@» °∆t° 15:59, 10 September 2024 (UTC)

Thanks. It's a brand new article so an old update to this module couldn't here be the reason, a link issue to Commons or so is of course perfectly possible. It would be nice if there is a way to make the error more meaningful, but no big deal if that is impossible (or too much work for the few cases it would be needed). Fram (talk) 16:02, 10 September 2024 (UTC)

Feature request: enable manual title-linking of open access stable identifiers in Cite book

{{Cite journal}} automatically links the title to a freely available external resource when it is marked with |ident-access=free; {{Cite book}} not only doesn't do this, but doesn't support manual title-linking to open access resources. There are plenty of open access books these days, and plenty of books with DOIs, free and subscription both. Can this functionality be enabled for this template? Could cut down on unnecessary URLs and their attendant archival and rotting. Folly Mox (talk) 14:24, 31 August 2024 (UTC)

The issue for book citations is that it's impossible to know if the chapter is meant to be linked, or the book title meant to be linked. And cite journal only links if |doi-access=free is set (or |pmc= given). It should also link when other identifiers of record (e.g. |jstor-access=free) are set. Headbomb {t · c · p · b} 15:17, 18 September 2024 (UTC)

Linking restrictions

Several parameters, e.g., |publisher=, allow wikilinks. The documentation should note that "|" must be escaped as {{!}} and that the pipe trick does not work, although normal piped links do. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 13:42, 10 September 2024 (UTC)

The documentation is not protected. If you believe that improvements can (should) be made, please do so,

—Trappist the monk (talk) 14:00, 10 September 2024 (UTC)

Now that I've seen the text that you added to Template:Citation Style documentation/publisher, I gotta ask why? Are you seeing some sort of problem with properly piped wikilinks in |publisher=? For me, both of these work:

{{cite book |title=Title |publisher=[[Random House|Random]]}}

Title. Random.

'"`UNIQ--templatestyles-0000001B-QINU`"'<cite class="citation book cs1">''Title''. [[Random House|Random]].</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=Title&rft.pub=Random&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1%2FArchive+96" class="Z3988"></span>

{{cite book |title=Title |publisher=[[Random House{{!}}Random]]}}

Title. Random.

'"`UNIQ--templatestyles-0000001F-QINU`"'<cite class="citation book cs1">''Title''. [[Random House|Random]].</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=Title&rft.pub=Random&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1%2FArchive+96" class="Z3988"></span>

Can you explain why you think it necessary to escape the pipe?

I suspect, though I haven't tried it, that the pipe trick can be made to work outside of <ref>...</ref> tags. But, since it won't work inside <ref>...</ref> tags, I don't think that we should bother to 'fix' it until MediaWiki fixes their end.

—Trappist the monk (talk) 16:55, 10 September 2024 (UTC)

I thought that I had seen it fail. And, yes, the pipe trick works outside of <ref>...</ref>. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 16:38, 12 September 2024 (UTC)

Line feed characters in quote within a citation

The article Electoral fraud in the United States had a really horrible citation added, which I removed in this diff. In the form I removed it, it didn't work because it ignored line feeds which were in both the wikitext and in the original document. I tried to make it work with {{cite report}} but couldn't figure out a way to make line feeds work. Any suggestions? Jc3s5h (talk) 20:00, 10 September 2024 (UTC)

That would be this thing?

Replace the linefeeds with {{pb}}. Rewriting it with {{pb}} we get this:

This also works in |quote=.:

But, if that quote is truly needed for the article (I don't think that it is – the source is free-to-read) then put it in <blockquote>...</blockquote> tags and cite it. Quotations require citations; citations do not require quotations.

—Trappist the monk (talk) 21:43, 10 September 2024 (UTC)

I worked out a version in my sandbox. In my version, I put a double quote at the beginning of each middle paragraph, which is the correct in a multi-paragraph quote that doesn't use a block quote. I also used the cite report template with as many parameters as I could find from the source. Jc3s5h (talk) 21:55, 10 September 2024 (UTC)

I agree it's a monstrosity. I mentioned this discussion in the talk page of the relevant article. I'll leave it to the proponent of the quote to find this discussion. Jc3s5h (talk) 21:56, 10 September 2024 (UTC)

100% agree it was way too much before Superb Owl (talk) 23:59, 10 September 2024 (UTC)

What about just the last paragraph and just for the parts not already in the text but could merit inclusion based on their coverage in reliable sources for easier context? Something like this Superb Owl (talk) 23:58, 10 September 2024 (UTC)

IMO this content should be summarized in the article, not quoted, and certainly not quoted with multiple paragraphs. Johnjbarton (talk) 01:12, 11 September 2024 (UTC)

FWIW the article has about 30 duplicate citations, the most I've ever seen in one article. If they were combined, the total number of citations could be reduced by about 10%. -- GreenC 04:52, 18 September 2024 (UTC)

I was unaware of the {{Duplicated citations}} template. Thanks for pointing it out. Jc3s5h (talk) 06:22, 18 September 2024 (UTC)

Thanks for flagging those - most are now fixed and we're now down to 355 citations (the other 5 duplicates had different quotes to help solidify an article in flux). Superb Owl (talk) 06:26, 18 September 2024 (UTC)

References

Lua error

Do not know what the problem is but getting "Lua error in Module:Citation/CS1/Configuration at line 2083: attempt to index a boolean value." with change to Dominic West when published. It shows OK when in preview without errors. Keith D (talk) 12:35, 19 September 2024 (UTC)

Appears to be an intermittent issue, as with the #Errors at line 2083 section above. As an immediate solution purging the article clears the error. A search for the error in articles turns up 258 articles, being a mix of already fixed articles and others needing purging. These are all new since the previous section. -- LCU ActivelyDisinterested «@» °∆t° 13:19, 19 September 2024 (UTC)

Spot checking the search results there's a lot more articles with errors than the last time. -- LCU ActivelyDisinterested «@» °∆t° 13:33, 19 September 2024 (UTC)

More generic titles

Hello, another generic title which is not very useful is |title=x.com about 450 of them at the moment. Keith D (talk) 17:14, 10 September 2024 (UTC)

Another one is |title=Amazon.co.uk about 45 around now. Keith D (talk) 19:38, 22 September 2024 (UTC)

Dissertation versus thesis in template

Is it possible to get Template:Cite thesis to say dissertation? If Template:Cite dissertation redirects to it, then it should at least support both. For now it marks up as: Text

You can override it by changing "type=" but that is just a workaround, and not built into the template. Why? I Ask (talk) 22:38, 21 September 2024 (UTC)

|degree= is a template-specific alias of |type=. How is a simple template to know that the source you are citing is a dissertation and not a thesis?

—Trappist the monk (talk) 22:46, 21 September 2024 (UTC)

Yes, but there is no option to change it through the visual editor. That seems like an issue. Why? I Ask (talk) 22:59, 21 September 2024 (UTC)

You did not say that your problem is with ve/templatedata. Your OP suggested that you wanted some sort of change to the {{cite thesis}} template or to Module:Citation/CS1 that underlies all of the cs1|2 templates.

VE uses Template:Cite thesis § TemplateData where someone has made |type= an alias of |degree=. I don't (won't) use ve so I did not know this until recently, but ve doesn't know about aliases even though the aliases are listed as aliases in TemplateData. Apparently the hack to work around that is to treat each alias as if it were not an alias. Per WP:SOFIXIT, here is your task: remove |type= as an alias of |degree=; add |type= as a proper parameter.

—Trappist the monk (talk) 23:17, 21 September 2024 (UTC)

We should add support for archive-access.

I propose we add support for an archive-access parameter. Why? I think https://en.wikipedia.org/wiki/Cardiac_stress_test#cite_note-13 would benefit from an indicator that a freely accessible copy is available at the archive by an archive-access=free parameter (which I've added for now, anticipating the support.) (many/most people will assume that because it's currently paywalled, they can't access it.) The green open lock would help a good fraction. It would also be useful on the two+ other pages that use the same source. RememberOrwell (talk) 10:13, 17 September 2024 (UTC)

This sort of thing is a FAQ. The answer has always been no because the purpose of an archive is not to skirt paywalls nor should we openly be making that our stated aim else these websites look at Wikipedia and remove their content from the Wayback Machine entirely which is trivially easy to do, and many have. If users figure this "feature" out on their own, more power to them. -- GreenC 23:26, 17 September 2024 (UTC)

Thanks. Can you point to where consensus was previously achieved? (Before posting, I searched for archive-access and found no discussion.) Actually, no need - It's been resolved; there's a similar parameter. archive-format=free. RememberOrwell (talk) 14:16, 18 September 2024 (UTC)

If an archived copy is not free to access, it should be removed. What are subscribers going to do? Log into the publisher's website via the archive snapshot? There's no need to tag an archive as free because they are assumed to be free, and should be. Folly Mox (talk) 14:37, 18 September 2024 (UTC)

Do not misuse cs1|2 parameters. |archive-format= (and any other format parameter) has a specific meaning: the file format of the linked source. See the documentation.

—Trappist the monk (talk) 14:46, 18 September 2024 (UTC)

Can you point to where consensus was previously achieved, @GreenC? RememberOrwell (talk) 05:27, 25 September 2024 (UTC)

I would add that our documentation is explicit: "As a courtesy to readers and other editors, editors should signal restrictions on access to material provided via the external links included in a citation." Adding support for an archive-access parameter is a logical way to do so. RememberOrwell (talk) 05:33, 25 September 2024 (UTC)

Category:CS1 errors: generic name

sports. 1234qwer1234qwer4 15:44, 24 September 2024 (UTC)

trans-parameters

I see "trans-title", "trans-work", "trans-website", ...; but there are missing parameters for "trans-author...", "trans-first...", "trans-last..." ; if the work's name is being translated, then the author should also appear in the original language form to match the source, and a translated version, to link to their Wikipedia article and show English rendering. That would prevent citation rot, if some editors transliterate names and that doesn't preserve source identifiablilty -- 64.229.88.34 (talk) 23:11, 22 September 2024 (UTC)

We had a related but not identical discussion at Help talk:Citation Style 1/Archive 92 § Proposed script-author parameter. Folly Mox (talk) 01:09, 23 September 2024 (UTC)

The |script-author= proposal doesn't seem to have been implemented -- 64.229.88.34 (talk) 01:30, 25 September 2024 (UTC)

Correct: it has not been implemented. We're still using |author-mask= to display author names in both native and transliterated forms. Folly Mox (talk) 22:31, 25 September 2024 (UTC)

National Register of Historic Places

In Streamliner there is a ref to National Register of Historic Places; is there a specific template or should I use {{cite web}}/{{cite document}} to cite a record?-- Carnby (talk) 19:31, 25 September 2024 (UTC)

There are several references on that article to NHRP documents, of which some seem to be primary sources that may not really have encyclopaedic value. (For example, I'm not sure what value is added by citing a facsimile of an NHRP application form for § Ships.)

For most other pages, {{cite web}} would work fine for these kinds of sources. {{Cite document}} is a sort of "last resort" citation template, for when you don't even have a URL and your source isn't a normal type of publication like a book or periodical.

Category:National Register of Historic Places templates has a surprisingly large number of member templates, none of which appears to be a bespoke citation template for the source, although it looks like Template:NRHP url can be embedded within a CS1 template to save typing / confuse bots. Folly Mox (talk) 22:26, 25 September 2024 (UTC)

@Folly Mox Thank you, how can I clean the page of these unnecessary references?-- Carnby (talk) 09:11, 26 September 2024 (UTC)

Unfortunately there's no way to determine which references are good and support claims cited to them apart from manual verification. I removed the single application form cited at Streamliner, and moved it to the ==External links== section of the main article MV Kalakala, where it was also not actually used as a reference. Then I gnomed all the citations there till I made myself late for work again. Folly Mox (talk) 11:35, 26 September 2024 (UTC)

How 'bout shorter PubMed URLs?

Should bots that are cleaning up CS1 replace URLs like https://pubmed.ncbi.nlm.nih.gov/18566177/ with URLs like https://pubmed.gov/18566177? I think so. (Annoyingly, URLs like https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4345631 can't be shortened to https://pubmed.gov/PMC4345631 because it doesn't redirect to the right location. "/articles" can be removed though.)

In other words, I propose https://pubmed.ncbi.nlm.nih.gov/<PMID>[/] be replaced with https://pubmed.gov/<PMID> when bots are doing other CS1 cleanup (same rules). RememberOrwell (talk) 10:34, 17 September 2024 (UTC)

Really these URLS should just be removed entirely. It makes it look like there's a full free version when there's not. No opinion on shortening the URL used by the template/identifiers though. Headbomb {t · c · p · b} 10:53, 17 September 2024 (UTC)

You should always use the direct link and not a redirect, even if that redirect is neater. As websites change overtime redirects are usually the first thing to be deleted or lost. Pubmed is likely more stable than most, but I would still apply the general rule. -- LCU ActivelyDisinterested «@» °∆t° 10:56, 17 September 2024 (UTC)

If a CS1 citation contains a valid full URL for a PMID, it should be converted to use |PMID=. That way, if the URL changes, the CS1 modules can be fixed and all of the citations will continue to work. – Jonesey95 (talk) 11:14, 17 September 2024 (UTC)

In practice websites do not always change all URLs equally. They change some, delete some, keep some. The top-down method is how silent linkrot is introduced. -- GreenC 23:22, 17 September 2024 (UTC)

Thanks @Michael Bednarek for restore.

PubMed.gov is a dedicated domain set up for this purpose, which suggests stability. Often a PubMed page does link to a free copy without a PMCID. Leads me to wonder if anyone has made a display tweak to surface sci-hub links. (Not arguing the main site should offer them by default, that would be a nonstarter, even for a free-content website. Heavy corporate presence...) RememberOrwell (talk) 14:26, 18 September 2024 (UTC)

Free copies tangent

PMIDs links never link to free copies. When there's a free copy, it's linked via PMCIDs. Bots handle that. Headbomb {t · c · p · b} 15:11, 18 September 2024 (UTC)

What are you talking about @Headbomb - that's frequently not true; example here with two links under "full text links". False - example in the OP above : #We should add support for archive-access. False - example in the OP above : #We should add support for archive-access. RememberOrwell (talk) 05:24, 25 September 2024 (UTC)

In both case, this is the DOI link, which can/will be automatically flagged as free-to-red by Citation bot with |doi-access=free, and the PMCID link, which will automatically link/be flagged as free to read by the template when making use of the |pmc= parameter. Putting the pmid url will suppress these free-to-read links in favour of the abstract-only PMID link. See

PMID in the url - links to abstract only

• Asim, A.; Kumar, A.; Muthuswamy, S.; Jain, S.; Agarwal, S. (2015). ""Down syndrome: An insight of the disease"". Journal of Biomedical Science. 22 (1): 41. doi:10.1186/s12929-015-0138-y. PMC 4464633. PMID 26062604.

No URL specified - autolinks to the full PMC version

• Asim, A.; Kumar, A.; Muthuswamy, S.; Jain, S.; Agarwal, S. (2015). ""Down syndrome: An insight of the disease"". Journal of Biomedical Science. 22 (1): 41. doi:10.1186/s12929-015-0138-y. PMC 4464633. PMID 26062604.

Headbomb {t · c · p · b} 05:34, 25 September 2024 (UTC)

OK, but there are better examples, and what I said is true: 1. PubMed.gov is a dedicated domain set up for this purpose, which suggests stability. 2. Often a PubMed page does link to a free copy without a PMC version. 3. Example in the OP above : #We should add support for archive-access disproves 2 earlier claims. CS1 could be more helpful at https://en.wikipedia.org/wiki/Cardiac_stress_test#cite_note-13 but agreed i- t could backfire to make them so, for GreenC's reason. I'mRememberOrwell (talk) 05:54, 25 September 2024 (UTC)

"Often a PubMed page does link to a free copy without a PMC version"

In which case it can be linked via |doi-access=, or a link to the actual free version. In the case of PMID 28799256, that's https://hal.science/hal-01688786. Headbomb {t · c · p · b} 07:53, 25 September 2024 (UTC)

Which has nothing to do with shorter PubMed URLs, right? Broken out. RememberOrwell (talk) 06:37, 27 September 2024 (UTC)

The point is whatever bot would 'shorten' these URLs should rather remove them so that they don't take the place of free links. Headbomb {t · c · p · b} 06:53, 27 September 2024 (UTC)

New linter category for duplicate IDs

In case you want something to fix, there is a new linter category for duplicate HTML IDs. It's Special:LintErrors/duplicate-ids. (Which may time out, here's the mainspace only version: [3].) Letting people on this page know because a lot of them are due to duplicate CITEREFs. Izno (talk) 20:45, 26 September 2024 (UTC)

Aren't duplicate CITEREFs to be expected? Articles using inline references could use different sources that have the same author and year, and the wiki software already deals with this automatically. -- LCU ActivelyDisinterested «@» °∆t° 21:31, 26 September 2024 (UTC)

Yes. It's still bad to emit duplicate IDs - the software has previously just pooped out the duplicate IDs, but it might not in the future, because they should be unique. I am not sure what we should advise in this case to fix the issue(s) particularly for cites that have this problem. Izno (talk) 21:33, 26 September 2024 (UTC)

What are the CITEREFs being used for other than short form references? The norm is inline references and will require no fixes. This seems to be an issue that should be dealt with my the software. -- LCU ActivelyDisinterested «@» °∆t° 21:39, 26 September 2024 (UTC)

What are the CITEREFs being used for other than short form references? Nothing to my knowledge.

This seems to be an issue that should be dealt with [by] the software. I mean, you can't have it both ways: either MediaWiki checks every page for duplicate IDs or it doesn't. It's very rare when MediaWiki adds hacks for user-generated constructs that are malformed in one or another ways, and especially not malformed HTML. As it happens, this issue was spawned by both the fact that the HTML gods say that thou shalt have only one ID on any specific web page, and by the fact that it was specifically short references where we didn't know which pages have duplicate IDs to fix. (I filed phab:T200517 in 2018 based on the discussion which made ref=harv the default.)

The solution is available to us today, |ref=none. I just want to see if anyone else wants to pipe up with another suggestion or other potential workaround.

Ignoring that, this may also identify duplicate inline references as well. Izno (talk) 21:53, 26 September 2024 (UTC)

Which seems like an argument for restoring |ref=harv and removing support for automatic CITEREF ID creation. The only difference from the past would be that all cs1|2 templates would needs do this including {{citation}} which as always had automatic CITEREF ID creation.

—Trappist the monk (talk) 21:51, 26 September 2024 (UTC)

That might be a valid approach after sub-referencing is available for a while. I am personally a fan of default ID generation and think opting out with |ref=none (or perhaps a new keyword, |ref=duplicate or |ref=duplicate-id) is the appropriate fix, not opting in via |ref=harv. I anticipate on most pages it is the rare case that has a duplicate ID generated from one of these templates (lint errors are per instance per page, not per page listings). Izno (talk) 21:57, 26 September 2024 (UTC)

Yah, |ref=none works and I would support |ref=duplicate.

—Trappist the monk (talk) 23:01, 26 September 2024 (UTC)

At the time that I looked at that report, there were 169k articles? pages? but all that you can see is one report page at a time. Who thought that was a good idea? cs1|2 uses <cite id="CITEREF...">...</cite> tags but, the report returns nothing when 'Filter by tag name' is set to cite. Sigh.

I can imagine that a bot might trawl the articles in the report looking for articles that don't use {{sfn}} or {{harv}} referencing and setting |ref=none for those templates that match the CITEREF ID in the report. But 50 at a time? Nah, I don't think that I'll be writing that bot – too much manual labor.

—Trappist the monk (talk) 23:01, 26 September 2024 (UTC)

I think the API allows more results, 50 is just what's presented in the UI, but I don't know who has tried to integrate with these reports. That's something that can be inquired about at WT:LINT. I'd suggest using Special:Search to get article lists but that doesn't have an integration (despite a long-requested one).

Filter by tag name I think is exclusive to certain other lint reports that actually check the tags (like "obsolete tag" filter). Izno (talk) 23:35, 26 September 2024 (UTC)

No, it's not 169k articles/pages, it's 169k instances of an offending error. So at maximum that's 169k divided by 2 pages, and at minimum 169k divided by 20, pages (because that's the maximum identified by the linter per page). I gather the linter is still growing its known pages at issue as I saw it at just 100k earlier. :)

As for such a bot, I would add the requirement that the references not be identical (well, at least within the ref tags). That should call for adjustment beyond ref=none IMO. Izno (talk) 23:38, 26 September 2024 (UTC)

IMO, the report should be fixed/tweaked to ignore anything with CITEREF as a starting string, or at least give the option to ignore those 'errors'. Truly problematic CITEREF duplication errors are tracking in Category:Harv and Sfn multiple-target errors. Headbomb {t · c · p · b} 00:15, 27 September 2024 (UTC)

Which is itself rife for false positives such that it requires a whole whitelist of its own. No, that doesn't actually track them all that well. Izno (talk) 00:54, 27 September 2024 (UTC)

You're thinking of Category:Harv and Sfn no-target errors, which has some false positives. The multiple targets error category is clean. Headbomb {t · c · p · b} 02:14, 27 September 2024 (UTC)

(edit conflict)

Not true. Category:Harv and Sfn multiple-target errors only tracks multiple target errors when Module:Footnotes is invoked by an {{sfn}} or {{harv}}-family template; and then, only those CITEREF IDs that are targeted by CITEREF links from an {{sfn}} or {{harv}}-family template or other [[#CITEREF...|...]]sort of link.

—Trappist the monk (talk) 01:06, 27 September 2024 (UTC)

FYI, Special:LintErrors/duplicate-ids is now timing out:

The maximum request time of 60 seconds was exceeded.

[7faafbc6-d49d-4f2a-bc1d-b2e50a9b8109] 2024-09-27 01:34:03: Fatal exception of type "Wikimedia\RequestTimeout\RequestTimeoutException"

FeRDNYC (talk) 01:39, 27 September 2024 (UTC)

Joy, the mainspace query is now falling over also, yes. Izno (talk) 05:18, 27 September 2024 (UTC)

At this point, I think we can safely ignore these errors when they are reported for CITEREF duplication. I created an example in my sandbox. When {{sfn}} is not used, there is no confusion created by duplicate IDs, as far as I can see. The footnotes link properly to the correct full citations. When sfn is used, we properly emit an error, since multiple citations are viable targets for the sfn template. We already have a system for seeing and resolving those problems.

I have been fixing Linter errors for six years, and sometimes the WMF developers, without consulting with the people who actually use their software, create "error" tracking pages (or categories, if we are lucky; don't get me started on these faux Linter "categories") that are less helpful than they should be. Some of them, like the "large table" Linter tracking, have been hidden after feedback from gnomes that there were too many false positives and non-errors in the list. Maybe the duplicate IDs will be able to go that way too. It is currently listed as "high priority" at Special:LintErrors, which doesn't make a lot of sense to me, but maybe I misunderstand the sudden urgency of this tracking. – Jonesey95 (talk) 21:07, 27 September 2024 (UTC)

Considering there were already over 3 million pages in the category on the list before the query started timing out, I'd say that "high priority" is definitely off the table, yeah. (Three MILLION!) FeRDNYC (talk) 21:59, 27 September 2024 (UTC)

(edit conflict)

If you were to add a couple of more references to your sandbox someplace that are not linked by {{sfn}}:

<ref>{{cite book |last=Greene |first=EB |date=2020 |chapter=Chapter 1 |title=Title |page=34}}</ref>

<ref>{{cite book |last=Greene |first=EB |date=2020 |chapter=Chapter 6 |title=Title |page=169}}</ref>

then preview the page and look at Parser profiling data → Lua logs [Show] you can see the logging information that Module:Footnotes posts when the page is published or previewed. That information might be improved to separately list anchor IDs that aren't linked from {{sfn/harv}} etc. When there are duplicates in that list, Module:Footnotes might add a category that at least identifies articles that have duplicate IDs not related to {{sfn/harv}}. The same caveats apply: Module:Footnotes is creating CITEREF identifiers from wikitext that it can see in the article so cs1|2 templates buried inside template wrappers aren't detected.

—Trappist the monk (talk) 22:20, 27 September 2024 (UTC)

Self published

For those times when something self-published (not by me) is used as a reference, how do we format the "publisher" field in the template? This is what I did. Trying to get rid of a few entries from here.  Mr.choppers | ✎  02:10, 28 September 2024 (UTC)

The solution is to remove the location, because the location is the location of the publisher. If there is no publisher, there is no location. Headbomb {t · c · p · b} 03:12, 28 September 2024 (UTC)

The tracking category says you can also set |publisher=none to suppress the message. -- LCU ActivelyDisinterested «@» °∆t° 10:52, 28 September 2024 (UTC)

"None" is useful. Plenty of sources have a location but no publisher, including the one I formatted above and here is another one. Thanks,  Mr.choppers | ✎  13:45, 28 September 2024 (UTC)

In citations, the location is the location of the publisher. If there are no publisher, there are no location to report. Headbomb {t · c · p · b} 19:51, 28 September 2024 (UTC)

False. In some very old books, we know the city that it was published in but not a publisher. An example from [4]: Girolamo Ruscelli (1566), Le imprese illustri, Venice. In that particular example archive.org [5] says that it was printed by the press of Francesco Rampazetto but you won't find that on the title page and a printer is not exactly the same as a publisher. Our article Girolamo Ruscelli cites the book with a location but no publisher. —David Eppstein (talk) 20:05, 28 September 2024 (UTC)

Is it not the case that self-published sources fail WP:RS, so why are bothering? 𝕁𝕄𝔽 (talk) 14:27, 28 September 2024 (UTC)

Not all self-published sources are RS, but some could be per WP:SPS. Other non-RS sources could be used for WP:ABOUTSELF statements. -- LCU ActivelyDisinterested «@» °∆t° 14:31, 28 September 2024 (UTC)

You can also just put "no pub." or "no publisher stated" or "self-published ebook" (etc) in the publisher field. jengod (talk) 22:09, 29 September 2024 (UTC)

Feature requests

• I would like to be able to click a box that adds the PD-inline text to the end of a ref: "This article incorporates text from this source, which is in the public domain." I find that adding this template helps ward off alarmed new page reviewers who have found that a chunk of text in an article has triggered the copyright bot (because, of course, it's in the public domain so has been published multiple places on the web).
• I would like to be able to have add to the {{free access}} icon to any ref. This is distinct from (albeit similar to) the open access movement. The point, from my perspective, is to help college students, other destitute researchers, and the techno-phobic know which resources are going to be readily accessible with a click or two, contra resources that require login/membership/money/time etc. Further research can then start with the free/easy stuff and then if they need a deeper dive, they can start figuring out scholarly databases etc.

Thank you for all you do!! jengod (talk) 22:17, 29 September 2024 (UTC)

You should be able to insert this information between the closing }} of the citation template and the </ref> tag. – Jonesey95 (talk) 03:40, 30 September 2024 (UTC)

I frequently do this very thing! But I have been informed that an excess of templates makes long pages sad, so I was thinking if it was incorporated into the cite template that could reduce...computer sadness? I very much don't know what I'm talking about but wanted to float the idea anyway. :) jengod (talk) 03:42, 30 September 2024 (UTC)

Sneaking in this explanatory link to Help:Template limits above my earlier wall of text. Folly Mox (talk) 10:47, 30 September 2024 (UTC)

It sounds like the first thing would require something like a |license= parameter, where a {{source-attribution}}, {{usgovpd}}, etc could be transcluded. I don't think this will actually reduce the number of template calls, and if you're looking to click a box to add something like this (inside or outside the CS1 template) the place for that request is unfortunately mw:Talk:VisualEditor, meta:Community Wishlist, or phab:.

For the second thing, is the end goal having every source annotated with a lock icon of some colour? Or every source accessible online annotated with a lock icon of some colour?

The main problem with this (apart from all the labour of marking up every citation anywhere that doesn't already have a lock icon) is that access varies so much by end user. Some sources are region locked, and we have no way of knowing where the end user is. Most college students will have access to some academic publishers through institutional subscriptions, and we don't want to drive them away from good sources they could access through their library portal with red padlocks. Websites change their paywall policies, previews are added to and removed from google books (which are also region dependent), books are added to and taken down from Internet Archive, etc.

It's a hard problem, and requires both a deeper knowledge of reader status than the website records and makes available for parsing, as well as a significantly higher degree of maintenance that would take a full time team reviewing regular web crawler reports.

Or is the idea just adding green padlocks to regular webpages? Sorry I'm sleepy. Folly Mox (talk) 10:44, 30 September 2024 (UTC)

What a thoughtful response! I hadn't thought about how much user situations would vary. I don't tag every web page or news article for sure but I would tag a public domain book on Google Books or Library of Congress or HathiTrust, just because I'm not sure everyone is as obsessed with public domain cutoff dates as we are. Another situation is that once in a blue moon I will find something valuable on NewspaperArchive.com rather than Newspapers.com. I find their UI absolutely confounding and have yet to figure out how to share free links to clippings, so sometimes out of what is essentially neurotic guilt, I will tag those reference(s) with {{closed access}}. The underlying overall goal just hews to the idea of a "free" encyclopedia--how can we help people get access to high quality information at little or no expense to them. We sort of take ourselves (and the massive growth of the Internet) for granted now but I am old and when we started there was very much the idea that we were saving people from having to buy an encyclopedia on CD-ROM from a tech conglomerate LOL.

No real action is needed, I suppose I'm just musing, but I always get jealous that there's a JSTOR-access-level field where you can toggle to "free" but that's not the case for other types of sources. Don't mind me and carry on! jengod (talk) 13:32, 30 September 2024 (UTC)

CNN paywall

So, CNN just put its site behind a soft paywall, which is unfortunate. Does that mean we now have to go through every article that currently cites CNN and tag them with |url-access=limited? Is that even feasible or advisable? Maybe a bot can handle that. InfiniteNexus (talk) 17:56, 1 October 2024 (UTC)

I would love it if we had bots that marked paywalled sources the way OAbot marks open access sources. Citoid doesn't have anything like this in its functionality, nor to my knowledge does Citation bot. As a result, almost all of our paywalled citations are unmarked.

Sites where the whole of the domain is behind a paywall would be a good target for bot editing, and might help achieve some of the aims articulated by jengod in the thread above regarding signalling to students and other destitute researchers which sources they should bother clicking through to.

This is probably the wrong venue, though. Adding all these parameters doesn't seem contentious, but would require so many edits that even an AWB editor would be advised to go through WP:BRFA. Folly Mox (talk) 18:14, 1 October 2024 (UTC)

False warning when URL is 'usurped'

Hi everyone! This isn't a big problem, but it's worth looking into. I noticed that Module:Citation/CS1 gives a maintenance message when a ref is tagged with | url-status=usurped. An example is ICD-11, ref 62.

Maintenance messages are not the same as errors. The former are green, the latter are red. Maintenance messages are usually not visible, unless you make them visible, like I have done in my common.css.

When I preview a page with a usurped-tagged source, it shows this warning at the top:

"Script warning: One or more (...) templates have maintenance messages; messages may be hidden (help)."

Also, with me, the usurped-tagged refs have this bit tagged at the end:

"CS1 maint: unfit URL (link)"

GreenC bot has been (correctly) marking thousands of URLs as usurped as part of an effort to combat the passive spamming of an Indonesian gambling syndicate that targets Wikipedia; see Wikipedia:Long-term abuse/Judi for more info. This shady organization re-registers expired domain names and turns them into spam sites.

The maintenance messages must be a bug in Module:Citation/CS1, because I see no reason why the script must throw an "unfit URL" warning when a ref is correctly marked as usurped. Again, this isn't a huge problem, because maintenance messages usually can't be seen. But it's still a bug, so it has to be fixed someday. Cheers, Manifestation (talk) 19:40, 25 September 2024 (UTC)

@Manifestation That's not a problem, and it's not a warning. It's the exact job of that tracking category. Read the introduction at Category:CS1 maint: unfit URL:

This hidden tracking category lists pages with CS1 citations that use |url-status=usurped or |url-status=unfit.

The keywords unfit and usurped are intended to identify original URLs that point to live sites that are inappropriate: spam, advertising, porn, etc.

A URL that returns a HTTP 404 error is not considered to be unfit and, in such cases, editors should set |url-status=dead.

CS1 and CS2 templates in pages listed in this category should be checked to ensure that the unfit and usurped keywords are correctly applied.

Only Module:Citation/CS1 should directly add pages to this category.

Pages with this condition are automatically placed in Category:CS1 maint: unfit URL.

Some hidden tracking categories are just that: Categories that track valid applications of certain templates in certain conditions. FeRDNYC (talk) 17:46, 26 September 2024 (UTC)

(I don't know where the "Script warning" message is coming from, that sounds like some userscript you may have installed. I don't see that, I just see the CS1 maint message attached to the citation in the reflist. But the message itself is not a warning; some script is warning you of the message's existence.) FeRDNYC (talk) 17:49, 26 September 2024 (UTC)

The script warning is a MediaWiki thing. Module:Citation/CS1 places messages about error and maintenance issues in the yellowish box at the top of a previewed page. MediaWiki prepends the 'Script warning:' text to each cs1|2 message. See example at Help:CS1 errors § Controlling error message display.

—Trappist the monk (talk) 18:18, 26 September 2024 (UTC)

Huh. Thanks. I wonder why I'm not seeing that? Is it displayed even when using the Visual Editor edit interface (in source mode)? ...Could be the userCSS I installed to unhide the hidden messages also hides that message, I'll have to check. FeRDNYC (talk) 18:24, 26 September 2024 (UTC)

No idea about anything to do with ve; I won't go anywhere near that thing. No doubt there are lurkers here who do use ve who might answer your question. Seems like a design flaw if a MediaWiki messaging system doesn't work with with MediaWiki's ve.

—Trappist the monk (talk) 19:04, 26 September 2024 (UTC)

(On an unrelated note, that ICD-11 article is festooned with offsite links (to https://icd.who.int/), which are normally not permitted in article bodies. Was some sort of exception decided on for those links, or do those need to be turned into citations?) FeRDNYC (talk) 17:56, 26 September 2024 (UTC)

(To answer my own question, those offsite links are all created using {{ICD10}} and {{ICD11}}, so I guess they are exceptions.) FeRDNYC (talk) 18:26, 26 September 2024 (UTC)

@FeRDNYC:
"it's not a warning."
The preview message says it is.

"It's the exact job of that tracking category."
That's fine. The tracking isn't the problem. Tracking is good. The problem is the maintenance message. These messages imply that something needs to be maintained, i.e. fixed, i.e. cleaned up. But there's nothing to maintain / fix / clean up here. The script warning is unjust.

"I don't know where the "Script warning" message is coming from, that sounds like some userscript you may have installed."
It's coming from the MediaWiki software. I do not have scripts installed (local, global).

@Trappist the monk:
That is correct.

- Manifestation (talk) 18:30, 26 September 2024 (UTC)

The cs1|2 templates cannot override MediaWiki-provided text so we're stuck with 'Script warning'.

—Trappist the monk (talk) 19:04, 26 September 2024 (UTC)

These messages imply that something needs to be maintained, i.e. fixed, i.e. cleaned up. But there's nothing to maintain / fix / clean up here. The script warning is unjust. Well, to your latter point, I'd agree; there's no reason to be 'warning' the user per se. But as @Trappist the monk says, we're stuck with the "Script warning" text, and the "warning" is simply that there are any maintenance messages, of any form, for any citations on the page.

Regarding the maintenance message itself, the Category intro text provides the "call to action" for why those messages are both displayed and tracked: CS1 and CS2 templates in pages listed in this category should be checked to ensure that the unfit and usurped keywords are correctly applied. Granted, that's a bit specious since there's still no way to remove a page from the category once it HAS been "checked to ensure...".

I suppose there's some argument to be made that |url-status=usurped / |url-status=unfit are (expected to be) infrequently-used parameters that warrant "keeping an eye on" them indefinitely. But, like I said, I understand your point that cleanup of that category is not possible, something that AFAICT makes it unique among the Category:CS1 maintenance subcategories.

There's a third type of tracking for the citation templates beyond error and maintenance: Category:CS1 properties. It mostly contains subcategories that don't generate visible messages and aren't meant for cleanup; maybe "unfit url" would be better tracked with a property instead of a maintenance message, @Trappist? (In fact, Category:CS1 properties even contains Category:CS1: abbreviated year range, a set of transclusions which do seem like they have cleanup potential — more so than unfit-url, anyway. So it feels like a bit of an inversion to have year-range-abbreviated in properties but unfit-url in maintenance.) FeRDNYC (talk) 01:29, 27 September 2024 (UTC)

FeRDNYC, the existence of a widely transcluded external link template doesn't necessarily mean it's always fine and good to put it in body prose (most transclusions of {{ICD10}}, for example, seem to be in navboxes). In this case though— given this is the article about the reference work the template points to, and there are like three hundred external links usually backy-back to an internal link, I'm really unconvinced putting them into citations or removing them would benefit anybody, irrespective of any MOS acronyms they contravene. Folly Mox (talk) 01:53, 27 September 2024 (UTC)

True enough, and an excellent point. In fact, the documentation for those templates says they're "primarily intended for use with {{medical resources}}" (an external-links formatting template). FeRDNYC (talk) 03:03, 27 September 2024 (UTC)

So... it this going to be fixed?

I like User:FeRDNYC's idea about turning this into a CS1 property. So I guess that would be Category:CS1 with a URL marked as usurped and Category:CS1 with a URL marked as unfit? Is there anyone who can implement this? Cheers, Manifestation (talk) 18:11, 2 October 2024 (UTC)

I'm coming in late to this conversation, but are you looking for Category:CS1 maint: unfit URL? Or are you suggesting that the maint category should actually live in Category:CS1 properties? – Jonesey95 (talk) 00:53, 3 October 2024 (UTC)

Please read the entire conversation. MediaWiki gives a false maintenance warning when a ref is marked with | url-status=usurped. Have a nice day, Manifestation (talk) 13:36, 4 October 2024 (UTC)

Cite podcast

Is it possible for someone to add "season" and "episode" to the Template:Cite podcast parameters? I do not feel comfortable editing such a widely used template myself. Why? I Ask (talk) 02:54, 2 October 2024 (UTC)

You'll be wanting {{cite serial}} if your podcast has seasons and episodes. Folly Mox (talk) 05:44, 2 October 2024 (UTC)

No, I do not want cite serial. That refers to a collection of episodes and seems to lack a parameter for specific seasons and episode number that something like cite episode has. Why? I Ask (talk) 11:35, 2 October 2024 (UTC)

My mistake, Why? I Ask. I misread the documentation. I also misremembered the conversation from last time this was idea was raised here, at Help talk:Citation Style 1/Archive 91 § Add new parameters for {{cite podcast}}, where it turns out I had suggested {{cite episode}} and nothing else happened.

Maybe it is time to add |series=, |episode=, |series-url=, |series-link= to {{cite podcast}}. Someone may have a better idea though. Folly Mox (talk) 12:20, 2 October 2024 (UTC)

language=en

I'll come clean: I'm pretty sure citations should literally never specify |language=en (incl. variants) and I strip it with a regex from the citations in every article I spend a while editing. The only possible edge case is when a source is in multiple languages, where |language=en,grc is perfectly informative and necessary. I guess I'm asking if there's something I'm missing here, and that there is some utility to this being specified sometimes? Remsense ‥ 论 09:53, 2 October 2024 (UTC)

I've had the same thought, but it comes in super handy when translating an article into another language. See for example Help talk:Citation Style 1/Archive 93 § url-status=dead (and a side-topic about language=en). Folly Mox (talk) 12:09, 2 October 2024 (UTC)

Automatic citation tools add it. This is probably why it's so common. I've included it this way but have never written it out by hand, Rjjⁱⁱⁱ (talk) 14:34, 2 October 2024 (UTC)

Cards on the table: I'm asking in the most totalistic manner possible whether anyone could be annoyed if I folded this regex into a userscript posted for general use. The only other thing I was able to think of was translation onto other language Wikipedias, but that seems beyond our remit, if luckily also totally solvable for those concerned. Remsense ‥ 论 22:34, 2 October 2024 (UTC)

I wouldn't support removing it en masse, if some editor finds it useful then it should be left. The same as with url-status or access-date. -- LCU ActivelyDisinterested «@» °∆t° 23:05, 2 October 2024 (UTC)

Of course it shouldn't be removed for its own sake, I'm speaking about one cleanup fix used with others as not to generate disruption, to be clear. But that's what I'm asking about: what possible use could it have? I don't think we need to assume something is potentially useful if no evidence or argument whatsoever has presented itself. Remsense ‥ 论 23:09, 2 October 2024 (UTC)

If some editors find it useful, then removing it is not a fix - it's personal preference. -- LCU ActivelyDisinterested «@» °∆t° 19:23, 3 October 2024 (UTC)

I broadly agree. Remsense ‥ 论 00:09, 4 October 2024 (UTC)

I wouldn't mind if anyone's automated script decided to remove |language=en in the course of other improvements. It's automated scripts that brought us most of these in the first place, so seems like fair play. Folly Mox (talk) 00:58, 3 October 2024 (UTC)

It has been established that this parameter combination is useful. It should not be removed. – Jonesey95 (talk) 17:42, 3 October 2024 (UTC)

It has also been established that this parameter combination is noisy boilerplate, to quote the discussion from last year that I linked above. I don't think it should be removed en masse wherever it's encountered, but parameters can be trimmed down when performing improvements to existing citations: full dates for book publications can be reduced to years only, access dates for physical media can be commented out, authors and editors past the first / fourth / seventh / whatever can be hidden, URLs duplicating stable identifiers can be deleted, archives of paywalled landing pages can be dropped, publishers can sometimes be removed, as can |via= if it matches the source domain, bibcodes and s2cids can be cut if editors find them more clutter than they're worth, and I think |language=en is another thing that can go on the list.

This isn't something I typically remove anymore (unless I'm completely doing over a citation from scratch), but it's a borderline case: some editors involved in tool-assisted translation find it useful, some editors who do citation work in the source editor find it bloaty clutter.

Anyway, |language=en is probably the thing on my list of personal preferences above that I care least about, but I did want to register one contrary opinion for the record. Folly Mox (talk) 00:39, 4 October 2024 (UTC)

If I remember correctly, |language=en used to cause Module:Citation/CS1 to emit a maintenance message and there were herds of gnomes who went about stomping-out that message. Allowing |language=en and then suppressing its display was prompted by the editors at Wikipedia:WikiProject Medicine/Translation task force to ease the translation of medical articles from English to some other language. Yeah, a small thing, but if it helps the translators, no burden on the rest of us. Translation is also why we prefer |language=en over |language=English.

—Trappist the monk (talk) 00:59, 4 October 2024 (UTC)

Here are a few links that I turned up with a rudimentary search. The search was not exhaustive; I am sure there are related discussions that are available via other search queries.

• Archive 30 (2017) of this page.
• Archive 8 (2015) of this page, where SMcCandlish described the current behavior of |language=en as reasonable.
• Module talk:Citation/CS1 (2015) in which multiple editors laid out multiple benefits, including copying references to other Wikipedias, simplifying the way that Citoid and CheckWiki work, and possible utility in research. Doc James specifically said I think it is good information to provide. Yes we are adding En references to nearly 100 other Wikipedias. This is the discussion to which Trappist refers immediately above, I believe.
• Talk:Tipping points in the climate system/GA1 (2022), in which Femke said I regularly remove these, but this comments has made me think about why these are included as default in automatic citations. I think it's for translations; reviewers have asked me to include this during a "FAC" on nlwiki if I recall correctly.

I'm sure there are more discussions out there that provide support for leaving |language=en in citations. – Jonesey95 (talk) 01:04, 4 October 2024 (UTC)

Thank you, I appreciate it. Remsense ‥ 论 01:18, 4 October 2024 (UTC)

For the record, my view on this has about-faced more recently that your 2015 prior discussion link. language=en is already the default for any source cited at en.wikipedia (and any other content here), so including this in citations is useless code bloat. Like the OP, I routinely remove this bloat when the value is en[-*], unless there's a special case like |language=en,gd. (Aside: I especially nuke any nonsense like |language=en-US, etc. We have zero need of any kind to try to identify the national variety of English the source is written in, since they're all mutually intelligible, the reader does not care, it serves no source identification/finding/verification purpose of any kind, and trying to do it in the first place is simply WP:OR, unless material in the source clearly says something like "I am choosing to write this article in Australian English" or whatever.) The comment above that |language=en is fairly common in citations simply because it is auto-added by some tools is correct; I have yet to find a single human editor who is doing this manually (if there is one, they need to knock it off). The tools doing this should not be doing it. (Aside 2: The tools should not be doing a bunch of other bad things that several of them are doing, like using |year= instead of |date=, putting website names in |publisher= instead of |work= AKA |website=, mis-coding author names as |author=Jane D. Smith instead of |first=Jane D.|last=Smith, or doing even more broken things like |author=Jane D. Smith & Philbert Gonzalez or |author=Mikael von Bek (Amsterdam Correspondent) or |author=Newsroom staff, etc.) A year or three ago, someone tried to argue here that including |language=en was somehow necessary for translation, but examination of this matter did not demonstrate that such a claim was true. If you are writing a script to convert English-Wikipedia citations to Russian ones (or whatever), you just default to the ru.WP template's equivalent of |language=en for any citation here that lacks a language code, just as someone on this site writing a script to convert from Russian citations will naturally assign |language=ru by default to Russian citations that lack the Russian citation template's equivalent of |language=ru. This is not rocket science.  — SMcCandlish ☏ ¢ 😼  05:31, 4 October 2024 (UTC)

I guess my final word is that, the parameter seems of marginal or no value for almost all editors, and while I'm glad it helps in translation I wish the responsibility of facilitating robust translation tools wasn't partially imposed on the wikitext that otherwise is intended to serve a fully different set of purposes. I trust that it happens to help though, so I don't want to raise anyone's hackles further about it. Remsense ‥ 论 05:36, 4 October 2024 (UTC)

'*' in archive-url timestamp

From Forensic accounting:

Godbole, Pradeep. "Fraud Investigation Methodologies and Report Structuring" (PDF). WIRC-ICAI. Retrieved 2024-09-23. {{cite web}}: |archive-url= is malformed: flag (help)CS1 maint: url-status (link)

The problem is https://web.archive.org/web/20230601000000*/ specifically the "*". The use of '*' in a timestamp is technically correct. There are at least 85 (search times out). I could eliminate them, they are not best practice, in most cases unnecessary, but there is a niche case if the citation concerns the Wayback Machine itself to show a range of snapshots to establish some fact. Looking through the existing cases I can't find any that are legitimate for that niche case. But they are also not broken URLs. How do we want to handle these should they produce an error and be tracked in Category:CS1 errors: archive-url? I don't think so because the URLs work (not malformed), and there might be a few legitimate use cases. -- GreenC 16:10, 2 October 2024 (UTC)

The use of a splat in the timestamp may be technically correct but what purpose does it serve in the 15th (subsecond) position? Does Wayback Machine support archive snapshots at subsecond intervals? cs1|2 rejects the splat because when it occurs within the fourteen characters of a timestamp, Wayback Machine links to a calendar page display that offers the reader multiple possible snapshots. That really isn't what we want. We want a link to a single snapshot that supports the text in our article.

As for the niche case if the citation concerns the Wayback Machine itself to show a range of snapshots to establish some fact, the url for that calendar display should be in |url=. No doubt another achiving service might archive the Wayback Machine calendar display, the url for which then goes in |archive-url=.

—Trappist the monk (talk) 18:14, 3 October 2024 (UTC)

The star in the 14th or 15th position sends the viewer to the index page: this 15 vs. that 14. One could make the Wayback link the primary |url= although this is done so often incorrectly (looking at reFill) that various tools and bots automatically move it to the archive-url, so it won't last very long in practice, short of some mechanism to flag the exception and cooperation by the tools and bots. In any case if it's a 15 character timestamp that is causing trouble, why not include star as one of the flag strings, like this flag, which doesn't trigger the 14 limit error. -- GreenC 23:40, 3 October 2024 (UTC)

It isn't the 14-digit timestamp length limit that cs1|2 rejects; its the splat appearing anywhere in the timestamp. This, for example, uses an 8-digit timestamp-with-splat; ends up at the same place as the 14-digit timestamp-with-splat:

{{Cite web |title=Fraud Investigation Methodologies and Report Structuring |url=https://wirc-icai.org/images/material/Investigation-Methodologies-Report-Structuring.pdf |url-status=deviated |archive-url=https://web.archive.org/web/20230601*/https://wirc-icai.org/images/material/Investigation-Methodologies-Report-Structuring.pdf |archive-date=2023-06-01}}

"Fraud Investigation Methodologies and Report Structuring" (PDF). {{cite web}}: |archive-url= is malformed: timestamp (help)CS1 maint: url-status (link)

The splat is an indicator of an editor too lazy to specify the exact snapshot that supports the text in our article.

why not include star as one of the flag strings? Because it isn't a flag string so should not be treated as one.

—Trappist the monk (talk) 00:32, 4 October 2024 (UTC)

Oh. I thought some editors occasionally want the index because they are citing a range of snapshot dates to support a fact, like, a website worked up to a certain date after which it stopped working (as an example). If our goal is to eliminate splats entirely, that's trivial to execute. -- GreenC 00:50, 4 October 2024 (UTC)

Editing documentation at Template:Cite web

Currently in the Usage section the blank Full parameter set in horizontal format uses "|last= |first= |author= |author-link=" rather than "|last1= |first1= |author1= |author-link1=".

While either last or last1 works, would it be better for consistency with the documentation of other cite templates like Template:Cite book and Template:Cite news where in the "Full parameter set in horizontal format" uses last1, etc? Or even to change over to using it on all the examples and blank parameters? 🌿MtBotany (talk) 22:24, 3 October 2024 (UTC)

The documentation is not protected. If you believe that the documentation can be improved, please do so.

—Trappist the monk (talk) 22:31, 3 October 2024 (UTC)

Thanks, I did notice that, but I am an "ask first" editor if something seems more critical than the usual thing I edit. The story about the new admin who deleted the front page is a cautionary tale that I've taken to mean just because I can, does not mean that I always should. I've updated the full horizontal one, but I'm going to see if anyone speaks up to object before I do any more than that. 🌿MtBotany (talk) 23:09, 3 October 2024 (UTC)

Markup for parameters?

@BlaueBlüte: I've been using {{para}} in my edits. I noticed that BlaueBlüte used {{code|{{!}}publisher{{=}}self-published}} (rendering as |publisher=self-published) rather than {{para|publisher|self-published}} (rendering as |publisher=self-published). Is there a reason to prefer the former? -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 10:31, 4 October 2024 (UTC)

Both render the same. {{para}} is basically a quicker version of doing what BlaueBlüte did. Headbomb {t · c · p · b} 10:48, 4 October 2024 (UTC)

Yes, but my question was whether there is a reason to not use the quicker version, or whether it is okay to continue using it. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 12:17, 4 October 2024 (UTC)

When things render the same, everyone is free to write things how they prefer. I can write Pokémon, or Pokémon. You won't know which way I wrote either unless you inspect the code. Headbomb {t · c · p · b} 12:23, 4 October 2024 (UTC)

@Chatul, feel free to change it if {{para}} is the more commonly used markup method in CS1 template documentations. There is something to be said for code consistency even where the rendered result is the same. ―BlaueBlüte (talk) 17:06, 4 October 2024 (UTC)

Should Cite journal allow both "pages=" and "page="?

Inclusion of both "pages=" and "page=" is considered to be an error, but it makes sense to allow both, giving the start and end page numbers in an article, and also the specific page on which some content appears. An example is <ref name= "Abrahamson1976">{{cite journal | last = Abrahamson | first = James L | year = 1976 | title = David Starr Jordan and American Antimilitarism | jstor = 40489774 | journal = [[The Pacific Northwest Quarterly]] | volume = 67 | issue = 2|url=https://www.jstor.org/stable/40489774| pages = 76–87|page=79}}</ref>, used in the article David Starr Jordan, where the content supported by the article is on page 79 of a 12-page article. Or is it better just to cite the page supporting the text? I have dealt with this in the article by appending "P. 79" to the citation giving the page range. But allowing both parameters is problematical if several pages in the article are required to support the text; maybe we don't need to cite the range, or to append the relevant pages after the template, so no change to the template required, maybe a note in the documentation. Best wishes, Pol098 (talk) 14:48, 5 October 2024 (UTC)

I would say it's almost always unneeded information, e.g. to specify both the page range of a chapter in a book or an article in a journal issue, as well as the pages being cited. What is allowed is both |page(s)= and |quote-page(s)=, where you can specify both the whole range being cited and the specific location of the direct quote you've also included in the citation. Remsense ‥ 论 15:02, 5 October 2024 (UTC)

That's nice theory. What is done in practice? What is done in the citation autofill?

Parameter names do matter. SamuelRiv (talk) 15:29, 5 October 2024 (UTC)

beg your pardon? Remsense ‥ 论 15:30, 5 October 2024 (UTC)

I think the comment by SamuelRiv means that the difference between "page" and "pages", while sometimes ignored, is important. Pol098 (talk) 21:42, 5 October 2024 (UTC)

Sorta, but not exactly. I am saying that particularly when editors use {{cite journal}}, they will fill |pages= with the article's full page range, as is done in typical print journal citations. The autofill template tends to do this too, last time I checked. That this behavior is not correct is not obvious from the parameter names -- in fact one would assume the template could easily automatically distinguish between a single page and multiple pages (and adjust "p." vs "pp." output accordingly), as CS1 is typically sophisticated in this regard.

Editors respond instinctively to parameter names, and the learn by existing examples on articles, so seeing both |page= and |pages=, what is a typical editor going to do? Ideally, we'd have statistics of 'improper' usage from citation correction bots.

But from experience, this is what happens in practice, and it's easy enough to address: replace the |pages= parameter with some other name, or else remove it, and have CS1 auto-detect plurals (as it already does in a passthru for detecting page range syntax). To address editors who really want to put in a bibliographic page range somewhere, have page-start (or plus page-end) parameters, and decide not to display them if you really want. SamuelRiv (talk) 22:09, 5 October 2024 (UTC)

In a nutshell, no. If you need to specify both, the format is |pages=76–87 [79]. Headbomb {t · c · p · b} 15:55, 5 October 2024 (UTC)

I've also done / seen |pages=1–85 at p. 59, |pages=39–71 (esp. pp. 41–42) etc. People overload the parameter a lot due to the inability to provide both a full page span and specific supporting information location (unless as noted a quote is also included in the template).

I know Citation bot used to overwrite specific page numbers with the full page span of the article / chapter being cited, but I think that may have been fixed last year. Automated tools might sometimes still do this if not programmed to avoid overwriting existing parameter values, but I don't keep up with all of them or use any of them.

It would probably be safer to be able to specify both a full page span and specific supporting page(s) so we don't have to rely on future coders realising that the |page= parameter their tools are modifying might be intentionally specific rather than an "error" since it doesn't match their database.

I would imagine the easiest implementation would be some alias for |quote-page= that doesn't depend on the presence of |quote=. No strong feelings really, since overloading the parameter works fine for our purposes. Folly Mox (talk) 18:18, 5 October 2024 (UTC)

Page numbers for journal cites should be the page range of the whole journal article, period. Any other information that specifies the location of the reference within the article should be specified elsewhere, either via a quote or outside the citation template. Be aware that the pagination of journal articles frequently differs from the pagination of their preprints and the pagination of their reprints (in collected works books and the like), and pagination may not be visible at all to readers of html online versions of journal articles. So when possible more specific locations should be described in other ways, for instance by section numbers and titles rather than pages. —David Eppstein (talk) 18:41, 5 October 2024 (UTC)

I hadn't considered those cases, thanks for cluing me in to why it can be clearly desirable. Remsense ‥ 论 21:10, 5 October 2024 (UTC)

What I find endlessly fascinating is how this is completely counter to the CS1 documentation.

(I know we've been told to edit the documentation, but as we see here, there is fundamental disagreement as to what actually the parameters do, disagreement in theory vs practice, and from some disagreement about whether practice should be considered.) SamuelRiv (talk) 22:16, 5 October 2024 (UTC)

Pol098, you can also use {{rp}} following the closing <ref> tag, like {{rp|79}}. {{rp}} is widely used and nearly as widely disliked. The WMF's future subreferencing extension may address situations like this. Folly Mox (talk) 19:14, 5 October 2024 (UTC)

{{rp}} is excellent, I strongly recommend it. DuncanHill (talk) 20:27, 5 October 2024 (UTC)

{{rp}} is a fucking plague of awfulness. Any other alternative is better. Headbomb {t · c · p · b} 20:42, 5 October 2024 (UTC)

I agree. It is an abomination. —David Eppstein (talk) 21:06, 5 October 2024 (UTC)

Interesting points in response to my suggestion. While I do use {{rp}} (sorry) I don't think it's ideal in this case. What I take away is that "pages=" can be overloaded without error; and that a specific page number can become incorrect with republication, etc. (as can a reference to a book not tied to a particular edition and publisher). Best wishes, Pol098 (talk) 21:39, 5 October 2024 (UTC)

Original title for a translated work

Given that there are dedicated parameters to note translators, is there also a way (apart from workarounds like the one I just used at Serbo-Croatian phonology) to note the original title of a translated work, and also the original language of the work? All the other translation-related parameters are about translations that Wikipedia editors themselves provide for works in languages other than (Standard) English, but I'm talking about works that are already present in translation (usually to English). --Florian Blaschke (talk) 22:28, 11 October 2024 (UTC)

I would usually provide complete citations for both the original work and the translation. Kanguole 22:36, 11 October 2024 (UTC)

 Courtesy link: Special:Diff/1250687271

I'll provide a complete citation for the work in original as well, if it feels important to note the original work in full.

Often I'll just swap the parameters |title= and |trans-title=, such that the original title appears in the square brackets following the English title chosen by the translators and their publisher. In these cases, I'll leave out the original language of the work, which usually doesn't matter. Folly Mox (talk) 16:54, 12 October 2024 (UTC)

I'm not convinced that using a parameter against its documented purpose is a good idea. The documentation for |trans-title= in {{cite book}} is here. If the English-language-titled source is the source that was consulted, putting the original-language title in |trans-title= just creates confusion and pointless work for later editors. If you consulted both the original and the translation, cite them both but cite them separately; don't attempt to mash two sources with (likely) wholly different bibliographic details into one template. The cs1|2 templates are not designed for that.

—Trappist the monk (talk) 18:33, 12 October 2024 (UTC)

Thank you for the clarification, Trappist. I'll discontinue the practice. Do you have a suggestion for how best to include this information if the original language version was not consulted (as appears to be the case here)? Folly Mox (talk) 20:06, 12 October 2024 (UTC)

§Further reading? But why? This is the English Wikipedia; listing a source that most of our readers won't be able to read and that wasn't consulted in the writing of the article strikes me as something akin to: omg!-there-is-an-empty-box!-I-must-fill-it! If the article is about the author or about this particular source then, sure, include the original language version. For articles on other topics? I don't see the need.

—Trappist the monk (talk) 21:20, 12 October 2024 (UTC)

The field "others" is not accepting the specified content

Greetings and felicitations. In Lore_Segal#Work there are nine citations for books which use "others = Illustrated by ___" as specified by the {{Cite book}} instructions, but the citations are still generating errors. Is there something wrong with the how the information was entered or with the instructions, or is there actually a bug? —DocWatson42 (talk) 23:28, 12 October 2024 (UTC)

Maintenance messages are not errors. |others=, by its very name, implies that there are 'other' contributors. cs1|2 templates are expected to be complete. This one, for example, is missing an author:

{{cite book|title = Tell Me a Mitzi|date = 1970|others = Illustrated by Harriet Pincus|lccn = 69014980|publisher = [[Farrar, Strauss and Giroux]]}}

Tell Me a Mitzi. Illustrated by Harriet Pincus. Farrar, Strauss and Giroux. 1970. LCCN 69014980.{{cite book}}: CS1 maint: others (link)

Presumably Segal is the author, but anyone reading that template in isolation wouldn't know that. If Segal is the author then write the template like this to include all of the necessary bibliograph information and still render as the broken template does (sans message):

{{cite book |last=Segal |first=Lore |display-authors=0 |title=Tell Me a Mitzi |date=1970 |others=Illustrated by Harriet Pincus |lccn=69014980 |publisher=[[Farrar, Strauss and Giroux]]}}

Tell Me a Mitzi. Illustrated by Harriet Pincus. Farrar, Strauss and Giroux. 1970. LCCN 69014980.

The above ensures that all of the necessary metadata are present:

'"`UNIQ--templatestyles-0000005D-QINU`"'<cite id="CITEREFSegal1970" class="citation book cs1">''Tell Me a Mitzi''. Illustrated by Harriet Pincus. [[Farrar, Strauss and Giroux]]. 1970. [[LCCN (identifier)|LCCN]]&nbsp;[https://lccn.loc.gov/69014980 69014980].</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=Tell+Me+a+Mitzi&rft.pub=Farrar%2C+Strauss+and+Giroux&rft.date=1970&rft_id=info%3Alccn%2F69014980&rft.aulast=Segal&rft.aufirst=Lore&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1%2FArchive+96" class="Z3988"></span>

—Trappist the monk (talk) 00:33, 13 October 2024 (UTC)

Ah—that's what I was missing. Thank you. ^_^ —DocWatson42 (talk) 08:09, 13 October 2024 (UTC)

Citing something jointly published by a work and a non-work

For {{Infobox US university ranking}}, this citation is to a publication that was made jointly by The Wall Street Journal (which should be italicized as a work) and College Pulse (which shouldn't take italics). Is there any way to use the |work= and |publisher= parameters to get The Wall Street Journal/College Pulse to display in the citation somehow? Sdkb ^talk 16:01, 15 October 2024 (UTC)

That sounds like a reasonable solution to me, or |publisher=College Pulse |via=Wall Street Journal Folly Mox (talk) 16:19, 15 October 2024 (UTC)

That sounds like a reasonable solution to me I didn't present a solution, so I'm not sure what you're referring to. The display I want to generate is the one in green quote text above. I don't think a standard |work= and |publisher= combo (or |publisher= and |via=) would be appropriate, since it would falsely imply that College Pulse is the publisher of the WSJ (and the via combo would imply that WSJ republishes content from College Pulse rather than jointly authoring it). Sdkb ^talk 16:43, 15 October 2024 (UTC)

I misunderstood. I erroneously thought you were suggesting |work=Wall Street Journal |publisher=College Pulse. True this would garbage up the metadata a bit, but seems like it would make sense visually. Seems like there are better ideas below. Folly Mox (talk) 18:01, 15 October 2024 (UTC)

Which one did you consult? Cite that one (WP:SAYWHEREYOUGOTIT). If you consulted both, cite both but cite them separately. cs1|2 is not designed to cite multiple sources with a single template.

—Trappist the monk (talk) 17:12, 15 October 2024 (UTC)

It seems from the cited website that "WSJ/College Pulse Rankings" is the title of a published series (often also published in other media, similar to the US News rankings and others), so would be good for a |work= field, with |title=2025 Best Colleges in the U.S., while WSJ in this case, because it's officially host on this medium, is probably the |publisher=. I'd try to use |via= only for a tertiary host. SamuelRiv (talk) 17:19, 15 October 2024 (UTC)

Using the publisher field would leave WSJ without italics. Sdkb ^talk 18:01, 15 October 2024 (UTC)

PMID value update

Currently getting an error trying to cite the following paper due to the PMID being out of range: https://pubmed.ncbi.nlm.nih.gov/39401844/

This help page says to report the issue here? Void if removed (talk) 13:53, 17 October 2024 (UTC)

Cite thesis

It might be good if this populated a hidden category hierarchy, with leaf nodes such as "Category:Pages including masters degree citations" populated by degree=masters|MSC|MBA etc. There should be a bucket for unrecognised types, and the system should be easy to extend.

The motivation is that in general we should not be using masters as citations (or indeed doctoral theses, though these are less of a red flag) instead using published works.

All the best: Rich Farmbrough 12:57, 18 October 2024 (UTC).

There's also Bacchelors' thesis / undergrad class thesis that we should keep track of if we start keeping track of those.

While we're on the topic, I'd like an error emitted if the degree type isn't specified. Headbomb {t · c · p · b} 13:09, 18 October 2024 (UTC)

Just out of curiosity, I repeated this search for |degree= and the first letter of the assigned value. Except for K, N, O, Q, and V–Z, every initial letter search returned results. There does not seem to be much consistency. I did not repeat these searches for |type=.

{{cite thesis}} is used in approximately 36,500 articles. Of those, ~11,000 use |degree= and ~15,200 use |type=. |type= and |degree= are partial aliases; |degree= causes Module:Citation/CS1 to append 'Thesis' to the value assigned to |degree=.

If we were to do do this we should probably have two properties categories; one for |degree= and one for |type=. Articles in the categories should be sorted by first character in the assigned value.

But should we? Is anyone going to actively do something with the data that are collected? Past experience says that nothing will will be done with these data.

Requiring {{cite thesis}} to have either |degree= (preferred) or |type= seems reasonable.

—Trappist the monk (talk) 13:51, 18 October 2024 (UTC)

Baccalaureate and undergrad theses should probably at least be tagged {{bsn}}, but maintenance categories for all the degree types sounds pretty unnecessary. Folly Mox (talk) 14:18, 18 October 2024 (UTC)

Adding {{bsn}} (or any other template) is outside Module:Citation/CS1's remit. Editors have not chosen to be even passably consistent with the value that they assign to |degree= so it is not possible for cs1|2 to categorize only baccalaureate and undergraduate theses; that is why I suggested sorting the categories by initial letter of the parameter's assigned value. And, no one here has suggested maintenance categories (yet).

—Trappist the monk (talk) 14:38, 18 October 2024 (UTC)

If you want an explicit category suggestion, I suppose it would be Category:CS1 error: degree missing or some such. Headbomb {t · c · p · b} 15:49, 18 October 2024 (UTC)

Trappist the monk, for sure, I didn't mean to suggest the module should add inline citation templates automatically. I was responding to the question about whether anyone going to actively do something with the data that are collected. Apologies for the miscommunication. Folly Mox (talk) 16:12, 18 October 2024 (UTC)

In all of this, let's remember that {{cite thesis}} and other cite templates are not always used to support claims in article text. Cite templates are often used in lists of works authored by a person, in which case a tag like {{bsn}} would not be applicable. – Jonesey95 (talk) 16:24, 18 October 2024 (UTC)

Best template to cite a webpage that is derived from a book

I noticed recently another editor citing one of my favorite webpages Flora of North America using the cite encyclopedia template rather than cite web as I have been doing. Would it be more correct to use cite book or cite encyclopedia instead of cite web for this source because it is first published in book form? 🌿MtBotany (talk) 16:34, 19 October 2024 (UTC)

In this particular case, I think citing it as a website is better, calling it e.g. "Flora of North America Online", because, as noted here, some of the webpages have been changed from the earlier print versions. Peter coxhead (talk) 17:03, 19 October 2024 (UTC)

Thanks Peter. I thought so, but better to ask than to guess. 🌿MtBotany (talk) 17:28, 19 October 2024 (UTC)

Title=none in newspaper and magazine citations

In {{cite journal}} (and {{citation}} with journal=), when there is no title url or link, it is acceptable to use title=none to suppress the title. This is useful in instances where there is no title or the title would be redundant, such as for book reviews where no real title exists (the review is just headed by a few lines of metadata about the item under review) or where several reviews are grouped together and would otherwise be given the same repeated and redundant title.

Less frequently, this would be useful to do on {{cite news}} and {{cite magazine}} as well. An example of this is Freshman's dream, where (to demonstrate the wide spread of the topic idea at the time of its original publication in 1938), one reference lists the original newspaper publication with its title and then four subsequent publications of the same piece, with the title deliberately stripped. In this case, the source code for Freshman's dream includes some very hacky code to hide the visible CS1 error messages that would be generated by omitting the titles, but it fails to suppress the error category that is also generated. It would be much cleaner if these could be formatted with title=none, but that doesn't work: for news or magazine citations (in this case as generated by {{citation}} and the newspaper= or magazine= parameter), this produces an actual visible "none" in the title position in the citation.

One alternative for Freshman's dream would be to pretend that these are all journals and use the journal citation ability to set title=none. A second alternative would be to give up on the citation templates as unable to handle anything non-formulaic and format the citations manually. A third alternative would be to modify the template code to be less inflexible and allow title=none for those citations. Or maybe there is a fourth alternative that I have missed. I don't think it's acceptable to force these citations to have titles, nor to try to extend the hacky code already there to also suppress the category. Are the third or fourth alternatives possible, or do we have to give up on using citation templates? —David Eppstein (talk) 01:11, 20 October 2024 (UTC)

David Eppstein, withholding comment on the substance of this matter, but I did suppress the error category using a non-hacky method (in addition to implementing a possibly dubious format change). Please feel free to revert if desired. Folly Mox (talk) 16:36, 20 October 2024 (UTC)

Thanks! So fourth alternative: no-tracking =y. We do still have the hacky error message removal code that it would be nice to avoid, but that's less urgent. —David Eppstein (talk) 16:40, 20 October 2024 (UTC)

DOI limit should be bumped

doi:10.70163/0036-0252.1157 is valid. The limit should be bumped to 10.80000 Headbomb {t · c · p · b} 19:05, 21 October 2024 (UTC)