Help talk:Citation Style 1/Archive 83

Vancouver
Hi. Question about Vancouver-style fields (vauthors/veditors): can you point me to documentation or demonstration about reformatting the parser output? I'm looking for reasons to avoid enclosing the whole content in double parentheses: "Special markup can be used to enforce that a value will nonetheless be accepted as written. The markup for this is, i.e., wrap the entire parameter value in two sets of parentheses." Thanks. fgnievinski (talk) 20:01, 25 October 2021 (UTC)


 * I have moved your question to a page where it may be answered. --Izno (talk) 19:34, 19 January 2022 (UTC)
 * , if you still have a question, an example citation, or a link to a page and a citation number about which you have questions, would help us answer your query. – Jonesey95 (talk) 00:09, 21 January 2022 (UTC)
 * this issue came up in the context of developing a citation style language file (CSL) using Wikipedia's citation template: . fgnievinski (talk) 03:49, 22 January 2022 (UTC)
 * Would it help to use author1 etc. with vanc? That way you can wrap individual authors with accept-this-as-written markup without sacrificing checks for the others. Hairy Dude (talk) 19:25, 2 March 2022 (UTC)

What's the right way to fix these references to prevent CS1 maintenance messages?
Hi everyone? What's the correct way to fix references like these that contain templates, so they don't generate the CS1 extra punctuation and multiple authors maintenance messages?

1) From Azmi Bishara, which uses the lrm template:



2) From Alec_Douglas-Home, which uses the long dash template:



3) From Flake (band), which uses the sic template:


 * Note: source has band named, Flakes.

4) From New Way (Jewish newspaper), which uses the lang template:



Thanks! GoingBatty (talk) 06:51, 14 February 2022 (UTC)


 * 1) I don't have any good ideas. The use of the template is for the benefit of editors, and is correct. What kind of message does it generate? One possible solution would be an invisible comment stating, "This citation contains bidirectional text". That is an ugly solution.
 * 2) Needs the author-mask instead of using the template
 * 3) Citations to catalogs (such as this one) are to be avoided. Citations should point to a source, not to an intermediate location. Unless the point of the citation is to prove that a certain work existed in that catalog. If that is the case (and note the catalog proposes a Wikipedia citation format) more editor info should be inserted in sic through the parameter Flake, and the reader info should be moved outside the citation.
 * 4) There is an incorrect use of version here. That input seems to belong to at. Also, no comma between vol # and vol title.
 * 172.254.222.178 (talk) 13:03, 14 February 2022 (UTC)
 * For most of these: don't write malformed cs1|2 templates; do not misuse cs1|2 parameters:
 * 1) there isn't really a best solution for rtl names preceding a wholly numerical date in the rendering. Two options are: write &lt;bdi>عزمي بشارة&lt;/bdi> or include the unicode U+200E character as the rightmost character in عزمي بشارة‎.  Both contaminate the metadata.  If the source has a romanization of the Arabic name, use that instead.
 * 2) do not write malformed citation templates. author is to hold the author's name.  When it is desirable to mask the name, two options: n or 0
 * 3) do not write malformed citation templates. do not use  in cs1|2 template parameters (see its documentation).  Flake (or Flakes) is not the author of the Trove catalog page so does not belong in author; Du Monde Records is not the publisher of the Trove catalog page so does not belong in publisher.  National Library of Australia is the publisher so via is misused.  Perhaps rewrite like this:
 * 4) do not write malformed citation templates. Too much detail is too much detail.  The encyclopedia entry title appears to be ПЕРИОДИ́ЧЕСКАЯ ПЕЧА́ТЬ so that is the text that belongs in entry.  Do not misuse version to hold something that is not a version indicator.  Do not use  or  templates in cs1|2 template parameters except quote.  Perhaps rewrite like this:
 * —Trappist the monk (talk) 14:15, 14 February 2022 (UTC)
 * 3) do not write malformed citation templates. do not use  in cs1|2 template parameters (see its documentation).  Flake (or Flakes) is not the author of the Trove catalog page so does not belong in author; Du Monde Records is not the publisher of the Trove catalog page so does not belong in publisher.  National Library of Australia is the publisher so via is misused.  Perhaps rewrite like this:
 * 4) do not write malformed citation templates. Too much detail is too much detail.  The encyclopedia entry title appears to be ПЕРИОДИ́ЧЕСКАЯ ПЕЧА́ТЬ so that is the text that belongs in entry.  Do not misuse version to hold something that is not a version indicator.  Do not use  or  templates in cs1|2 template parameters except quote.  Perhaps rewrite like this:
 * —Trappist the monk (talk) 14:15, 14 February 2022 (UTC)
 * 4) do not write malformed citation templates. Too much detail is too much detail.  The encyclopedia entry title appears to be ПЕРИОДИ́ЧЕСКАЯ ПЕЧА́ТЬ so that is the text that belongs in entry.  Do not misuse version to hold something that is not a version indicator.  Do not use  or  templates in cs1|2 template parameters except quote.  Perhaps rewrite like this:
 * —Trappist the monk (talk) 14:15, 14 February 2022 (UTC)
 * —Trappist the monk (talk) 14:15, 14 February 2022 (UTC)
 * —Trappist the monk (talk) 14:15, 14 February 2022 (UTC)


 * GoingBatty, in general, I would completely disregard what is stated above about metadata. Any metadata transmission should respect the citation data, not the other way around. There is no policy or guideline indicating that citation metadata are to be allowed to dictate what citation data would be, neither is this a burden of citations editors.
 * As for #3, it has to be made clear that sources have to be cited directly, not as catalog items. A catalog record may be cited when the record itself is considered the source referenced by wikitext. In that case, Trappist's citation is correct. In the particular article, the reference is supposed to support the following wikitext: In the context of the article, this may not be cited as a catalog record. It is better to have a footnote along the lines of "See release information at the Trove catalog, National Library of Australia : (insert link to catalog record)."
 * 65.88.88.126 (talk) 15:47, 14 February 2022 (UTC)

Thanks to everyone for their input. My bot has gone through Category:CS1 maint: extra punctuation and fixed all the low hanging fruit. Your help to fix the rest would be appreciated. GoingBatty (talk) 20:51, 27 February 2022 (UTC)

"Redundant" identifiers?
Is there any benefit to having, say, a pmid or s2cid in a citation when the only links they have to a source are the same as what the doi is? Citation Bot added a few of these to an article where the preexisting citation style was not to make use of these identifiers. To me, they just (inconsistently) clutter up the citations with extra links that don't provide the reader with any additional ways of accessing the source. Ditto adding a 10.2307/[JSTOR ID] when there already is jstor going to the exact same location. Curious if there's any sort of policy that says "if there is a pmid or s2cid, a citation should use it even if it provides no further assistance for a reader locating a source" or if WP:CITEVAR can encompass things like "this is not a journal about a medical topic so pmid won't be used". Thanks. Umimmak (talk) 09:42, 24 February 2022 (UTC)
 * The examples you gave certainly look like redundancies. Multiple content identifiers are useful when they lead to different content hosts and if not, obviously clutter up the citation. I suspect the problem is the bot's writer(s) did not take such cases into account. A bot can only do what the writer codes. 71.247.146.98 (talk) 12:31, 24 February 2022 (UTC)
 * There is no such policy.
 * Having all identifiers on the table may help some ad hoc user to find it at a local library. Even though we provide the links, remember that the text itself can be valuable for finding the source of interest. Izno (talk) 16:55, 24 February 2022 (UTC)
 * Not to mention that "|doi=10.2307/[JSTOR ID] when there already is |jstor=" often point to different locations as journals are purchased and sold. &#32; Headbomb {t · c · p · b} 04:22, 25 February 2022 (UTC)
 * I'm not convinced of the value of s2cid, but including any/all of bibcode/pmid/doi/jstor when a paper has them is definitely useful for readers who have easier access through one type of id than through another. (I tend to see papers that are on both Taylor & Francis and JSTOR fairly often, for instance, and have an easier time accessing the JSTOR versions than through the doi when that happens.) —David Eppstein (talk) 05:34, 25 February 2022 (UTC)
 * have easier access through one type of id than through another -- to clarify I'm only talking about when the doi goes directly to JSTOR. I think it's fine for a citation to have (ex. 1) (goes to OUP) and, but I don't really see the benefit to having (ex. 2)  (goes to JSTOR) and  or (ex. 3) , which has a link to the full text... but it's at SpringerLink which is where  goes. It's examples (2) and (3) which feel like clutter to me -- I don't see the benefit of the Citation Bot adding these and which I'm tempted to remove them. But if an identifier provides another way for someone to access the text, that I have no issue with. Umimmak (talk) 05:44, 25 February 2022 (UTC)
 * Where the doi goes to is not permanent. It may not even be the same location for different people who access the doi at the same time. That's the point of doi's: the resource can move to a different url and the doi should follow it, and there is supposedly a mechanism for directing people to the most appropriate copy of the resource for them. —David Eppstein (talk) 05:47, 25 February 2022 (UTC)

Self-transclusion, again
That the usage of these templates causes the page in question to transclude itself has come up at least twice before:


 * Help_talk:Citation_Style_1/Archive_55
 * Help talk:Citation Style 1/Archive 64

This is maddening behavior when working on unused template cleanup, because any otherwise unused template that includes a citation template on its page (Template:Glasgow–Aberdeen line is one example) will be excluded from standard reports. I understand the reasoning and good intentions behind this behavior but I really think that this has to be considered a bug. Mackensen (talk) 01:06, 25 February 2022 (UTC)
 * Do you have a suggestion for how to implement the article-level date format templates without transcluding the article in this way, so that the citation templates can see the date format template? If there is no workaround, the bug is in the limitations of the wiki software, not in the template. —David Eppstein (talk) 01:46, 25 February 2022 (UTC)
 * There's an important distinction between a limitation of the mediawiki software (no inherent page properties) and a bug (any page using a particular type of template unexpectedly transcludes itself). Maybe this feature is important enough to justify that outcome, though I don't think (from reading past discussions) that people realized it would do that. Offhand, if I were doing this anew I'd consider proposing a Wikidata property (e.g. "topic's date format") and having the module look at that instead. Wikidata already tracks other bits of article meta, like related categories. I can see all kinds of semantic objections to that approach, not the least that MOS:DATERET creates all kinds of problems with standardization. Still, it would work, and we'd trade a technical hack for a procedural one. Mackensen (talk) 02:43, 25 February 2022 (UTC)
 * Topic's date format would be rejected because it was single-wiki specific, most-likely.
 * There is a future on en.wp where there are similar separate fields for page-wide meta-data. Until then, adjusting the query of interest such that single use self-transclusions are reported separately seems like a good idea to me. Alternatively, since we know this particular module is both widely used and the most likely to do so, separate out low use templates that include a transclusion of Module:Citation/CS1.
 * Just some thoughts. I'm pretty sure the query system can handle that. Izno (talk) 03:28, 25 February 2022 (UTC)
 * I'd rather see article date format become a reader preference, in the long term, instead of having to encode it in any kind of structured data on a per-article basis. —David Eppstein (talk) 04:09, 25 February 2022 (UTC)
 * This is just a shot in the dark, but would it be possible to use or adapt Module:Excerpt, or some portion of it, for this purpose? If a page is configured correctly, use dmy dates should be in the lead section, which is where Excerpt looks by default. – Jonesey95 (talk) 06:50, 25 February 2022 (UTC)
 * Excerpt fundamentally uses the same API as we do, they just end up displaying the content instead. Izno (talk) 08:39, 25 February 2022 (UTC)
 * Auto date formatting was a user preference and subsequently removed by our request because it has negative parsing performance for users logged in and simple inconsistency for users logged out (because users logged in didn't know things were inconsistent for users logged out). So I don't expect that to happen. Izno (talk) 08:14, 25 February 2022 (UTC)

No url warning when doi is present
At Volkswagen emissions scandal a new reference was added with https://doi.org/10.1093/restud/rdac007. I naturally changed this to and also added subscription. To my surprise, the template showed a warning about the url-access field requiring a url, so I had to put the url field back in. The doi field can always be automatically translated to a url (the template even does this), so shouldn't the doi field count as having a url field?  Stepho  talk 23:03, 25 February 2022 (UTC)
 * With the exception of isbn and issn, every identifier is automatically translated to a url. Most identifiers link to a source that lies beyond a paywall.  Because of that, no identifiers support subscription (or   or  ).  URL-holding parameters, like url, are presumed to be free-to-read unless marked otherwise so, for them, free is not allowed.  Documentation for this is at.
 * —Trappist the monk (talk) 23:36, 25 February 2022 (UTC)
 * To be more specific: if you change a url to a doi, you should also change the url-access to doi-access. —David Eppstein (talk) 23:40, 25 February 2022 (UTC)
 * Which for this case is not necessary because the example doi points to a source beyond a paywall. free does not apply and subscription is not allowed.
 * —Trappist the monk (talk) 23:50, 25 February 2022 (UTC)
 * Thank you for pointing out doi-access. That means my choices are:
 * Use subscription - not allowed !
 * Do not use url-access and/or doi-access- doesn't tell the user about needing a subscription.
 * Use subscription - requires duplicating the unchanging doi with a url which can change at the website owner's whim - surely this is against the purpose of using the doi.
 * Of course I can use the last option (use both doi and url) but that does seem to go against the grain. Is there a reason why subscription is actively disallowed ?  Stepho  talk 05:46, 26 February 2022 (UTC)
 * Because we don't highlight the norm. Most doi identifiers lie beyond a paywall.  That is the norm.  Most urls are free-to-read.  That is the norm.  If we highlight the norm, medical- and scientific-article reference sections will bleed red; similarly, general interest article reference sections will bleed green – just so much visual clutter.  The access icons highlight source-links that are not the norm: a free-to-read doi-linked source; a url-linked source behind a paywall.
 * —Trappist the monk (talk) 13:07, 26 February 2022 (UTC)

Hanging/orphaned quotation marks
Noticed one today in a citation with quote where the line broke at the opening mark. Can a word joiner (U+2060) be used to rectify this? 65.88.88.126 (talk) 17:38, 28 February 2022 (UTC)
 * Where? In simple experiments I am not able to duplicate your claim: opening quote mark always stayed adjacent to the first letter of the text when I resized the window to force line breaks.
 * —Trappist the monk (talk) 18:30, 28 February 2022 (UTC)
 * Sorry, I know it is tricky to reproduce. It may also be a browser artifact, it was there in Edge (Windows 10) and Tor (Ubuntu 20), but not in Safari (Big Sur 11.6.4). Since these are separate machines with different monitors, I will try to fix the pixel width required for an example and update this. 65.88.88.62 (talk) 20:24, 28 February 2022 (UTC)
 * It could. However, this is a normal artifact of browsing. We shouldn't try to control wrapping here much less in general. (This comes up regularly between here and WP:VPT.) Izno (talk) 21:09, 28 February 2022 (UTC)


 * Ok, it is not a general case, my mistake. It happens when interp is used before the actual quote in the data field (pls adjust the window to see result, happens in all browsers/OS referred to above. Examples padded with extra fields to ease resizing).






 * Maybe specific to the interaction with that template. E.g. vanchor doesn't break.






 * (Demonstrate VisibleAnchor)


 * Oh well. 65.88.88.62 (talk) 21:13, 28 February 2022 (UTC)


 * Sheesh. It's the bracket...






 * Tried it with !( and html encoding, still breaks. Abandon ship? or fix is visible in the far horizon? 65.88.88.62 (talk) 21:32, 28 February 2022 (UTC)


 * Btw, WJ works, but I was looking for a programmatic solution:






 * 65.88.88.62 (talk) 21:37, 28 February 2022 (UTC)
 * Thanks for the examples. The quotation mark stays with all of the example left-bracket characters in my browser, Firefox for Mac. File a bug against your browser, I guess. Quotation marks should be nowrapped with the following character by the HTML renderer, IMO. – Jonesey95 (talk) 23:58, 28 February 2022 (UTC)
 * Ok. Just installed Tor 11 (built on Firefox 91.6) on a Mac 11 and the problem disappeared. It is still there for both Safari & Edge on the Mac 11, and Edge on Win 10. Will update Tor on the Linux machine and see if it resolves the issue. Thanks for pointing it out. 65.88.88.46 (talk) 15:48, 2 March 2022 (UTC)

Multiple quotes of source
It appears impossible to add more than one quote per cite template right now; can it (of course it can technically) be made possible please? 05:55, 2 March 2022 (UTC)
 * Please provide a real-world example in which someone would need more than one quotation to support a given claim in an article. If you really need to provide two quotations, you may need two citations. Also, you are free to add material after the close of the cite template but before the closing ref tag if necessary. – Jonesey95 (talk) 06:20, 2 March 2022 (UTC)
 * I'm thinking of the efficiency of the reuse of named refs for e.g. an interview which supports multiple claims. Multiple refs is the obvious alternative, but can lead to clutter. I will get back to you with a real-world example later. Cheers.  07:07, 2 March 2022 (UTC)
 * On second thoughts; I don't see the value of providing a real-world example other than to provide a narrow lane within which to argue. There are millions of possible cases of potential usage and each should be evaluated on a per case basis, therefore hypothetical evaluation of the potential pros and cons should suffice. Example: a single interview may cover multiple subjects covered by an article; currently if quote is used, the citation cannot be sensibly reused and, instead, the same source must be separately cited in support of other article claims. Having the ability add multiple quotes (as with authors e.g. quote1, quote2 and so on) would allow the same source to be reused, reducing clutter and making clear to readers that the same interview is cited multiple times for various article claims (as is the current state of affairs when quote isn't employed (sensibly at least)). If quote has any credible value, it seems like a no-brainer to me that multiple quotes should be possible per citation.  14:30, 2 March 2022 (UTC)
 * A good old ellipsis works for most simple cases. If you're quoting significant passages in this parameter, you should probably rethink that. Izno (talk) 07:44, 2 March 2022 (UTC)
 * An ellipsis to truncate a single quote is fine when the quote is entirely about one subject, but if a single source covers multiple subjects, it would be wildly improper to separate the various subjects with ellipses. Example: An interview with a pop singer who also breeds race horses; interviewer asks "What was your last single about?" to which the singer responds "... sunshine and cookies ...". Then later in the same interview the interviewer asks "Why are all your horses so short?" to which the breeder responds "... selective breeding ... short horses are cute ...". The resulting citation's quote would read "... sunshine and cookies ... selective breeding ... short horses are cute ...". If that would be acceptable, I'd argue quote should probably be deprecated and eventually removed as it's likely to be more confusing than useful.
 * I don't think I've ever actually used quote (if I have; hardly ever if at all) so don't have anything to rethink at this juncture.  14:30, 2 March 2022 (UTC)
 * Oh, no; I have used quote a few times for The Partisan.  14:33, 2 March 2022 (UTC)
 * A possible solution could be to use an sfn template for each quote - see Template:Sfn. GoingBatty (talk) 14:35, 2 March 2022 (UTC)
 * Noted (pun intended). Thanks.  14:50, 2 March 2022 (UTC)
 * Better to use the other examples there and avoid simply because dumping quoted text into a parameter designed to name a location, loc, is semantically incorrect.  See.
 * —Trappist the monk (talk) 14:57, 2 March 2022 (UTC)
 * If a quote is to an article, put the quote in the article body (or in a separate 'quotes' section) and cite it.  A single interview quoted multiple times for different purposes in an article needs only a single citation.  I think that quote should never have been invented because it clutters reference sections and is too often abused; enumeration of that parameter would multiply the clutter and opportunity for abuse.
 * —Trappist the monk (talk) 14:57, 2 March 2022 (UTC)
 * I would use block-level quote markup rather a special section to bundle all quotes. I agree that may be abused, and unless brief, uglifies (new word?) the citation. I try to only use the parameter to quote interesting/disambiguating aspects of the citation itself, sort of a stand-in  parameter. 65.88.88.46 (talk) 15:42, 2 March 2022 (UTC)
 * A Quotes section is an interesting idea.  15:09, 2 March 2022 (UTC)
 * No, not really. There will be the inevitable tendency towards adding more quotes than subtracting, possibly resulting in quote farms. Also, any such quotes would need to be given context, which involves more complex attribution/footnoting than an inline context-based quote would need. 65.88.88.46 (talk) 15:35, 2 March 2022 (UTC)
 * I envisaged a section like Notes with complete quotations from sources that are not suitable for inclusion inline, such that the citation can be applied to the statements in the article body and the quote, allowing the reader to see the specific part of the source cited right there on the article (the entire purpose of quote; correct me if I'm wrong) and as such, the context exists in the article body. I think a quote farm isn't a nice idea.  16:41, 2 March 2022 (UTC)
 * Is the use of quote ever valid (insofar that it solves a problem only it can solve)?  16:41, 2 March 2022 (UTC)
 * It is a proper question, as Trappist discussed above, but it is quite embedded into this citation system to change now, I believe. Perhaps more emphatic guidance can be given to include quotes in short citations, or in full citations only when necessary in context. Programmatically, I would suggest limiting the character-size of the field (gently) so that quotes in full citations remain brief and do not distract/overwhelm the citation itself. As noted above I believe there are justified uses relating to quotes in the citation itself. For instance, to point out an alternate spelling in the source by quoting it; to point out a source characteristic such as (previously) "classified" etc. if the source is so marked; to disambiguate a contributor role if again this can be quoted from the source, for example to point out that the cited editor in a journal/book is the editor of a particular department or article. And so on. 65.88.88.46 (talk) 17:23, 2 March 2022 (UTC)
 * I'm not convinced it's even a proper question. It conflates "valid use" with "only way to solve a problem", and sets an impossible hurdle (there is no citation template parameter that solves a problem only it can solve, because it is always possible to format citations as plain wikimarkup without citations). —David Eppstein (talk) 17:30, 2 March 2022 (UTC)
 * I'm not convinced it's even a proper question. It conflates "valid use" with "only way to solve a problem", and sets an impossible hurdle (there is no citation template parameter that solves a problem only it can solve, because it is always possible to format citations as plain wikimarkup without citations). —David Eppstein (talk) 17:30, 2 March 2022 (UTC)

Multiple urls of source
I couldn't resist riffing off the previous section heading. In Special:Diff/1075110483, I added two cite web templates. It's a two-part article for which I wanted to include the URLs for both parts. What I really wanted to do is a single citation but with two links for the two parts. Is that possible? -- RoySmith (talk) 13:42, 4 March 2022 (UTC)
 * No. cs1|2 templates are not designed to cite more than one source at a time.  You can always write:
 * (part 2)
 * That looks ok, but the part 2 url does not make it into the citation's metadata. You can spoof  and do this:
 * That method also looks ok, but in this case, neither url makes it into the metadata.
 * It is best to do it as you have done.
 * —Trappist the monk (talk) 14:14, 4 March 2022 (UTC)
 * Or use short ref templates, and add both in-source locations in loc:
 * 69.203.140.37 (talk) 18:53, 4 March 2022 (UTC)
 * It is best to do it as you have done.
 * —Trappist the monk (talk) 14:14, 4 March 2022 (UTC)
 * Or use short ref templates, and add both in-source locations in loc:
 * 69.203.140.37 (talk) 18:53, 4 March 2022 (UTC)
 * 69.203.140.37 (talk) 18:53, 4 March 2022 (UTC)
 * 69.203.140.37 (talk) 18:53, 4 March 2022 (UTC)

Online news sources - clarity between "work=" and "publisher="
I think it may be appropriate to note in the documentation that news agencies or publishers that primarily publish online or distribute news, such as CNN, NBCNews.com, Associated Press, etc. should use the "publisher=" parameter instead of the "work=" parameter. This should also apply to state organizations such as BBC News. However, news sites that are primarily associated with a newspaper, magazine or journal should use "work=", e.g. The Times, Los Angeles Times, TIME.com, etc. This would only put it in italics as far as I can tell. Thoughts? Facts707 (talk) 14:10, 5 March 2022 (UTC)
 * Sigh. This discussion again.
 * No. We cite the published, not the organizations that publish the work.  An organization's online presence is their 'published work' even when that online presence has the same name as the corporate entity that owns it.
 * cs1|2 templates create COinS metadata which allows the rendered citations to be machine readable. COinS does not support   (publisher) for 'periodical' templates .  Because of that, 'periodical' templates that use publisher when they should be using work (or an alias), produce incomplete metadata.  That important bit of information is not available to those readers who consume our citations via the metadata.
 * There are editors who will disagree with everything that I have just written... and so it goes.
 * —Trappist the monk (talk) 15:00, 5 March 2022 (UTC)
 * No (agree with Trappist). A related question: why is it so hard for people to realize that work and publisher refer to entirely different properties? They cannot be substituted for each other even when their names are identical. If Mr. X publishes a continuing autobiographical series titled "Mr. X" that has certain things in every issue, Mr. X the person is not "Mr. X" the autobiography. Don't approach Mr. X to find out what is in his bio, it's too complicated etc. Get the bio itself. 65.88.88.71 (talk) 16:30, 5 March 2022 (UTC)
 * Turning it around, why is it so hard for people like Trappist to realize that CNN, Associated Press, etc. are actually the correct names of organizations that publish things, and that it is a valid choice to put the name of an organization that publishes things in the publisher field instead of putting the thing they publish (which may or may not have exactly that name) into a different field? —David Eppstein (talk) 20:44, 5 March 2022 (UTC)
 * Well I guess for the reason stated about COinS, and doing it one way because standards have many benefits. Speaking as a programmer, who consumes this metadata at scale and tries to automate matching of records elsewhere (library catalogs) with cites on Wikipedia, it's useful to have less complexity and standards, as far as possible, I realize this is cat herding and the burden ultimately falls on application developers like myself to untangle the data. -- Green  C  02:17, 6 March 2022 (UTC)
 * ? I cannot speak for Trappist or anyone, but what keeps an editor from entering the publisher info? There is no suggestion that one should not use publisher. The problem arises when editors do not use work or its aliases because they mistakenly give publisher preference. A citation must provide the work where the proving info exists. For cosmetic reasons, the guidance suggests that if the work name is the same or almost identical to the publisher name, then the publisher info (as a secondary citation item) may be omitted. And again, nobody stops an editor from using it anyway, aesthetics be damned. 65.88.88.201 (talk) 15:55, 6 March 2022 (UTC)
 * 65.88: Template:Cite_web states "Do not use the publisher parameter for the name of a work (e.g. a website, book, encyclopedia, newspaper, magazine, journal, etc.).... Not normally used for periodicals. Omit where the publisher's name is substantially the same as the name of the work (for example, The New York Times Co. publishes The New York Times newspaper, so there is no reason to name the publisher)." GoingBatty (talk) 17:20, 6 March 2022 (UTC)
 * Agreed, but I believe what you point out (editors using publisher name for work name) may be exactly because of confusion when the names are identical or very similar. And perhaps also because they don't read the documentation properly. But notice the OP: publishers that primarily publish online or distribute news, such as CNN, NBCNews.com, Associated Press, etc. should use the "publisher=" parameter instead of the "work=" parameter. This is plain wrong. 65.88.88.201 (talk) 17:42, 6 March 2022 (UTC)
 * ? I cannot speak for Trappist or anyone, but what keeps an editor from entering the publisher info? There is no suggestion that one should not use publisher. The problem arises when editors do not use work or its aliases because they mistakenly give publisher preference. A citation must provide the work where the proving info exists. For cosmetic reasons, the guidance suggests that if the work name is the same or almost identical to the publisher name, then the publisher info (as a secondary citation item) may be omitted. And again, nobody stops an editor from using it anyway, aesthetics be damned. 65.88.88.201 (talk) 15:55, 6 March 2022 (UTC)
 * 65.88: Template:Cite_web states "Do not use the publisher parameter for the name of a work (e.g. a website, book, encyclopedia, newspaper, magazine, journal, etc.).... Not normally used for periodicals. Omit where the publisher's name is substantially the same as the name of the work (for example, The New York Times Co. publishes The New York Times newspaper, so there is no reason to name the publisher)." GoingBatty (talk) 17:20, 6 March 2022 (UTC)
 * Agreed, but I believe what you point out (editors using publisher name for work name) may be exactly because of confusion when the names are identical or very similar. And perhaps also because they don't read the documentation properly. But notice the OP: publishers that primarily publish online or distribute news, such as CNN, NBCNews.com, Associated Press, etc. should use the "publisher=" parameter instead of the "work=" parameter. This is plain wrong. 65.88.88.201 (talk) 17:42, 6 March 2022 (UTC)

Encourage linking of "work" and "publisher" in all occurrences in references?
In the new age of a substantial amount of news coming from worldwide online sources, it may be appropriate to encourage linking the work or publisher in all occurrences in citations to allow readers perusing the "References" section of an article to quickly find more information on the source of the news. Since citations only appear in References, this shouldn't be excessive (as noted in MOS:REPEATLINK) for the reader. Thoughts? Facts707 (talk) 14:18, 5 March 2022 (UTC)
 * That seems a 'article-level' recommendation so, to me, is out-of-scope for cs1|2. Perhaps this topic is better raised at WT:CITE.
 * —Trappist the monk (talk) 15:04, 5 March 2022 (UTC)

Why no extlink in department= ?
Today, I ran across a publication I wanted to cite, where I wanted to cite this using citation/cite journal, with title for the title of the contribution, department for the title of the department, journal for the title of the journal, and with a link to the online copy going onto the department parameter (also with a doi pointing to the individual contribution). However, the citation templates not only do not provide a way to put a link there, they go out of their way to prevent a link being put there. There is no department-url parameter, where I would expect a link to go. And in the absence of a parameter, the obvious workaround is to put a url in the department parameter itself, but the citation templates specifically look for any links there and flag them as errors rather than passing them through to the formatted output.
 * It is an individually authored contribution within a larger ongoing department of a periodical. Whenever this department appears in its periodical, it consists of multiple separate contributions.
 * Each contribution within the department has its own authors, and the department has an editor, who is not the editor of the whole periodical.
 * For the contribution I wanted to cite, I could find a url for an online copy of the whole department from the issue it appeared in, on the personal web site of the editor of the department.
 * I did not find a freely available online copy of the individual contribution I wanted to cite, separated from its department. It does have a paywalled copy, linked from a doi at the publisher's web site.

(The periodical in question is an academic newsletter rather than a peer-reviewed journal or popular-press magazine, and we have no specific citation format for that class of periodical, but that's a separate issue; citing it as a journal works well enough.)

Why is this an error? How can I put a link on the department of a publication, going to an online copy of the department that I want to link to? —David Eppstein (talk) 20:38, 5 March 2022 (UTC)
 * Hi there! If you could please provide the URL, it may help people to provide a suggestion for you.  GoingBatty (talk) 22:48, 5 March 2022 (UTC)
 * Url for department: https://www.cs.umd.edu/users/gasarch/bookrev/45-3.pdf Please note, however, I am not looking for suggestions on how to format this specific reference. I am looking for answers to my question on why the citation templates appear to have been deliberately made so inflexible. —David Eppstein (talk) 22:58, 5 March 2022 (UTC)
 * Because virtually nothing is done to the cs1|2 module suite without it gets talked about here, there was a discussion:.
 * —Trappist the monk (talk) 23:18, 5 March 2022 (UTC)

Using cite web or cite book or whatever for an ebook...
What is considered a WP Best Practice for when one must cite an ebook? I read Citing sources and all it says is
 * If there are no page numbers, whether in ebooks or print materials, then you can use other means of identifying the relevant section of a lengthy work, such as the chapter number or the section title.

But what do folks do when the only copy of a book available to an editor - I am trying to improve the Zapruder film article, it is in really sad shape, maybe a "C" - trying to use an ebook...how do you actually, really fill out a cite web? It is kind of hard to get specific for verifiability purposes... Thanks, Shearonink (talk) 17:39, 6 March 2022 (UTC)
 * An ebook is a book published in one of many available media. Use cite book. If the book is online, add the proper parameters to that template (url, etc.). 65.88.88.201 (talk) 17:46, 6 March 2022 (UTC)
 * I know about cite web but I am asking what is considered best practice and what are fellow editors' boots on the ground/actual editing practices...how can we render an ebook when there are no pages, how can we get as specific as possible. Do people ever just use quotes along with the chapter/section title? Actual experiences, y'know? Thanks, Shearonink (talk) 18:13, 6 March 2022 (UTC)
 * If you use a section/chapter title in chapter you have already quoted from the source. In the highly unlikely situation where a book-length work is not sectioned you can instead use At "quote from source" or y At quote from source. 68.173.76.118 (talk) 00:37, 7 March 2022 (UTC)


 * A book online is still a book so :
 * —Trappist the monk (talk) 17:51, 6 March 2022 (UTC)
 * You don't include a quote to anchor the reference? I always feel like I should include a quote or something... Shearonink (talk) 18:13, 6 March 2022 (UTC)
 * I don't because I think that quotes in a citation are an abomination. If the quote is important to the en.wiki article, put the quote in the en.wiki article and cite it.
 * —Trappist the monk (talk) 18:31, 6 March 2022 (UTC)
 * Yes, you will usually need a suitable in-work location. If a quote is the best you have, a quote is reasonable. Izno (talk) 18:32, 6 March 2022 (UTC)
 * I don't because I think that quotes in a citation are an abomination. If the quote is important to the en.wiki article, put the quote in the en.wiki article and cite it.
 * —Trappist the monk (talk) 18:31, 6 March 2022 (UTC)
 * Yes, you will usually need a suitable in-work location. If a quote is the best you have, a quote is reasonable. Izno (talk) 18:32, 6 March 2022 (UTC)

Help wanted
Sally Ride is giving me a "; messages may be hidden (help)." warning, but I have no idea what the problem is. Could someone have a look at it for me? Hawkeye7  (discuss)  20:48, 7 March 2022 (UTC)
 * . It was an extra comma in a citation author name. – Jonesey95 (talk) 20:52, 7 March 2022 (UTC)
 * Sally_Ride. Too many commas, Names only, don't include degrees, rank, title or other extraneous stuff.
 * To show maintenance messages, see
 * —Trappist the monk (talk) 21:03, 7 March 2022 (UTC)
 * I tried that, but it didn't work for me. Hawkeye7   (discuss)  00:30, 8 March 2022 (UTC)
 * It is not obvious from the history at User:Hawkeye7/common.css that you ever added the css to show maintenance messages. That css is this:
 * —Trappist the monk (talk) 00:54, 8 March 2022 (UTC)
 * Oh. I see. I enabled instead of . Another editor fixed the article for me.  Hawkeye7   (discuss)  02:18, 8 March 2022 (UTC)
 * —Trappist the monk (talk) 00:54, 8 March 2022 (UTC)
 * Oh. I see. I enabled instead of . Another editor fixed the article for me.  Hawkeye7   (discuss)  02:18, 8 March 2022 (UTC)
 * Oh. I see. I enabled instead of . Another editor fixed the article for me.  Hawkeye7   (discuss)  02:18, 8 March 2022 (UTC)

There are now Dozens and Dozens of Mainstream News Apps, Why no specific Citation Template for them?
This is maddening. News apps do not provide URLs, yet the Wikipedia citation templates penalize you for not including a URL-- with unattractive red bold-faced error messages in any citation added without a URL.

Wikipedia needs to quickly develop a citation that accommodates news apps, as they are very widely used. It mystifies me as to why this hasn't been done yet.

And also-- why has no reference citation that includes a space for detailing the specific "News app" been completed? Millions of people now use news apps.

Chesapeake77 (talk) 13:25, 8 March 2022 (UTC)
 * CS1 citation templates cite works, not media. Use cite news. All stories in such apps resolve to a URL. Have ever tried sharing one? News apps are no different from any other delivery format. If you want to make a specific-source template for a particular news app, you can do so. 74.64.150.19 (talk) 13:45, 8 March 2022 (UTC)

Transclusion of articles to themselves
On Danish Wikipedia someone asked at da:Wikipedia:Teknisk_forum why articles are transcluded to themselves. Example the article about Nordea where Nordea is listed under "Transcluded templates". Someone believe it is the CS1|2 module that is the reason. Is it supposed to be like that? --MGA73 (talk) 16:38, 8 March 2022 (UTC)
 * Yes, at line 592 in da:Modul:Citation/CS1/Configuration is this:
 * local content = mw.title.getCurrentTitle:getContent or '';
 * counts as a transclusion. See mw:Extension:Scribunto/Lua reference manual.
 * —Trappist the monk (talk) 16:52, 8 March 2022 (UTC)

Checking for date format on template pages
Related to the above section, is it necessary for the date format to be set at Module:Citation/CS1/Configuration when the current page is a template and not an article page which transcludes it? The reason I'm asking is that using  causes the page to transclude itself, which when it happens on templates, prevents the templates from appearing on unused reports if they are unused. Since the format of a reference in the template itself is dependent on where it is being transcluded, it seems that it really is unnecessary to check for the date format here. I propose adding a namespace check to see and not continue if it is on the template namespace. Gonnym (talk) 18:54, 8 March 2022 (UTC)


 * Seems directly relevant to . Adding a check for template namespace seems like a reasonable solution since auto-formatting is only really necessary for mainspace/draftspace. Izno (talk) 19:28, 8 March 2022 (UTC)
 * The templates are used in template space or so say these searches:
 * → ~150
 * → ~170
 * Do those templates use the templates to instruct editors in the desired format?  Are the  templates present so that cs1|2 will auto-format the dates in any included citations?  Are the  templates present in those templates for some other reason?
 * Where they are used in template space, is it ok to disable auto-date formatting?
 * —Trappist the monk (talk) 19:56, 8 March 2022 (UTC)
 * A template should not be linked to but transcluded, when it is transcluded then the date format of the transcluding article should be used over any date format present in the template page. There really isn't much to gain from having these there. Additionally, the small usage size (~320) compared the issues it is causing, is more of a sign that these should not be handled by the module. Gonnym (talk) 20:02, 8 March 2022 (UTC)
 * I think that I agree with what you just wrote but I'm having trouble understanding A template should not be linked to but transcluded and how that applies to what I wrote.
 * cs1|2 reads the article's unparsed wikitext when looking for a template.  The wikitext of transcluded templates does not appear in the article's unparsed wikitext.  Any  template that is transcluded into a template that is itself transcluded into an article is not visible to cs1|2.
 * —Trappist the monk (talk) 20:45, 8 March 2022 (UTC)
 * Agreed with Gonnym. Additionally, the number of interest is actually the intersection, 69 use dmy and 48 use mdy. And I'm finding that most of your search's set are entirely unnecessary. Izno (talk) 20:31, 8 March 2022 (UTC)
 * Yeah, my search is flawed, and yeah the templates that I looked at may have just been using the template to make the template's documentation page look pretty, and yeah, putting in a namespace restriction is going to spoil that.
 * —Trappist the monk (talk) 20:45, 8 March 2022 (UTC)
 * Sandbox tweaked so that  not called and   is not set when the 'article' is in template namespace.
 * —Trappist the monk (talk) 22:33, 8 March 2022 (UTC)
 * —Trappist the monk (talk) 20:45, 8 March 2022 (UTC)
 * Sandbox tweaked so that  not called and   is not set when the 'article' is in template namespace.
 * —Trappist the monk (talk) 22:33, 8 March 2022 (UTC)

Bug in Citation/CS1/Date validation
I am currently localizing the CS1 module for Estonian Wikipedia and found a bug. In Module:Citation/CS1/Date validation, in function reformatter:

If the date has two years, as is the case for 'My-My', 'dMy-dMy', 'Mdy-Mdy' formats, it erroneously replaces the first year (y) with the anchor year (a). What you want to do instead is to replace the second year (y2). This works for me:

Kaniivel (talk) 18:49, 12 March 2022 (UTC)
 * Thanks for that. I've implemented a fix in the sandbox:


 * —Trappist the monk (talk) 20:29, 12 March 2022 (UTC)

cite episode complains about formatting errors that dont exist
For some reason, when defining |format in a cite episode, itll produce the error "|format= requires |url=", even when the |url parameter is defined. Also, the format doesnt even appear anywhere. Is this meant to be disabled for this citation format or something? (Originally found this while editing Evanna Lynch, check reference 12).

Example: (<-- Error in question - url IS defined)

If anyone knows about this, please tell me why or if its intended. Thanks. Aidan9382 (talk) 18:27, 14 March 2022 (UTC)
 * When the new lua versions of the templates were created, the goal was to make them render in exactly the same form as their wikitext predecessors. Because the old  did not support format, the new  did not support format.  The new  template was created nearly seven years ago (18 April 2015).  I don't recall anyone complaining about format not working in  between then and now.  Fixed in the sandbox:


 * —Trappist the monk (talk) 19:27, 14 March 2022 (UTC)

Unrecognized languages
Would it be possible to set up a Lua table as a final fallback for language checking (or if the table exists already, point me to it) so that we can clear out Category:CS1 maint: unrecognized language (at least the mainspace ones)? I would expect to provide the IETF/ISO code of interest as the key (i.e., support the key only as input in the wikitext of a citation) and the name of the language as the value, so that in the future if MediaWiki comes to support a certain code, we will not be using the wrong name (unless deliberately set in the relevant table that exists as an override). Izno (talk) 03:31, 14 February 2022 (UTC)

The ISO 639-1 code for Irish is supposed to be “ga” but if I use the parameter “lang=ga” in a citation, it shows as “(in Ga).” Using the ISO 639-2 and -3 code of “gle” also fails. Xenophore ; talk 00:47, 16 March 2022 (UTC)
 * That happens because  is also a language name.  I'll think about how best to address this.
 * —Trappist the monk (talk) 01:02, 16 March 2022 (UTC)
 * In the meanwhile, this works:
 * —Trappist the monk (talk) 01:07, 16 March 2022 (UTC)
 * —Trappist the monk (talk) 01:07, 16 March 2022 (UTC)
 * —Trappist the monk (talk) 01:07, 16 March 2022 (UTC)

Error with url-status=live if no archive-url
The documentation says. However, if  is included in a Citation template, the error message   is displayed if the mouse is hovered over the citation number, in the "This is only a preview" box, and sometimes in the list of references. I have WP:boldly added a note to that effect in the documentation (it was quickly reverted); either the documentation must be changed, or the template must be changed to allow  without error message.

I suppose this could conceivably be an artefact of my particular setup, but it definitely happens.

Best wishes, Pol098 (talk) 15:18, 15 March 2022 (UTC)
 * What you wrote was this:
 * and that is not correct. It is possible to get a similar message:
 * cs1|2 templates emit a lot of maintenance messages. I don't know if it is worth the effort to note each message in the documentation of every parameter where that message might apply, but if we do, then the documentation must be correct.
 * —Trappist the monk (talk) 15:32, 15 March 2022 (UTC)
 * cs1|2 templates emit a lot of maintenance messages. I don't know if it is worth the effort to note each message in the documentation of every parameter where that message might apply, but if we do, then the documentation must be correct.
 * —Trappist the monk (talk) 15:32, 15 March 2022 (UTC)
 * —Trappist the monk (talk) 15:32, 15 March 2022 (UTC)


 * Thanks for response. As I've drawn attention to this issue amongst those who know the system, I won't do anything more. My opinion is that  should be removed from the   documentation, as it generates the warning when the mouse is hovered over a ref number in read mode, and as there is absolutely no point in adding   to a reference without an archive-url. But I'm not going to do anything or to argue, whatever is done. Best wishes, Pol098 (talk) 15:48, 15 March 2022 (UTC)
 * Note that the red messages are errors, while the green messages are maintenance messages. GoingBatty (talk) 21:26, 16 March 2022 (UTC)
 * Thanks, that's of course true. I should have made the distinction in my comment. I find it a bit disconcerting to see a maintenance message when hovering over a ref number in read mode, and seeing that maintenance warnings are reported when previewing an edit. But I have nothing else to say. Best wishes, Pol098 (talk) 22:08, 16 March 2022 (UTC)

archive-url and identifier-created urls
When url is omitted or empty, title in will be linked from the value assigned to pmc or, alternatively, doi (requires free). These identifier-created urls are not url so when access-date is present, cs1|2 emits an error message:

That same rule should apply to archive-url which also requires url but, in the live module, it doesn't. Fixed in the sandbox.

—Trappist the monk (talk) 19:39, 16 March 2022 (UTC)
 * I tend to agree that using access-date and archive-url for publications accessed in ways other than through their url is a mistake, so in philosophical terms I don't disagree with this change. But I am curious about two points. (1) We still support access-date when there is a nonempty chapter-url, contribution-url, or other url parameter, not just the main url= parameter, right? Will there be any way to supply an archive-url for those parameters, after this change? (2) Do you have any data on how many articles this will cause errors in? —David Eppstein (talk) 20:22, 16 March 2022 (UTC)
 * chapter-url, conference-url, contribution-url, entry-url, map-url, section-url, transcript-url are not supported by . Similarly, these url-holding parameters are not supported by  when journal is set.
 * I don't know how many templates do not have url but do have doi and free; or do have pmc; and do have archive-url.  Relatively easy to find templates with something; not so easy to find templates without something.  I only discovered this because I found a template that had pmc and archive-url where the archived url was the url in lay-url (which has never been supported by archive-url) but wasn't showing the archive-url-requires-url error message.  I doubt that that example is the only one of its kind but I cannot say how many more are out there.
 * —Trappist the monk (talk) 21:37, 16 March 2022 (UTC)
 * They may not be supported by cite journal, but they are supported by cite book and cite conference, which also allow references to have auto-linked doi but no url. I was assuming that you would want those other citation templates to remain consistent with cite journal. —David Eppstein (talk) 21:41, 16 March 2022 (UTC)
 * Auto-linking title from pmc or from doi (when free) is only supported by and by  (when journal is set).  Here are two  without url; one with pmc and the other with doi and free:
 * and the same for :
 * None of the four examples above auto-link title. I included journal just in case, but the result is the same if your take it away.  Can you give examples of any cs1|2 templates other than the cases I've demonstrated that auto-link title?
 * —Trappist the monk (talk) 22:02, 16 March 2022 (UTC)
 * I take this answer to be a denial of my assumption that you would want those other citation templates to remain consistent with cite journal. —David Eppstein (talk) 22:35, 16 March 2022 (UTC)
 * My answer was an attempt to show that your statement:
 * They may not be supported by cite journal, but they are supported by cite book and cite conference, which also allow references to have auto-linked doi but no url.
 * is inaccurate. If that is not what I have accomplished then I have failed to communicate.
 * —Trappist the monk (talk) 23:21, 16 March 2022 (UTC)
 * None of the four examples above auto-link title. I included journal just in case, but the result is the same if your take it away.  Can you give examples of any cs1|2 templates other than the cases I've demonstrated that auto-link title?
 * —Trappist the monk (talk) 22:02, 16 March 2022 (UTC)
 * I take this answer to be a denial of my assumption that you would want those other citation templates to remain consistent with cite journal. —David Eppstein (talk) 22:35, 16 March 2022 (UTC)
 * My answer was an attempt to show that your statement:
 * They may not be supported by cite journal, but they are supported by cite book and cite conference, which also allow references to have auto-linked doi but no url.
 * is inaccurate. If that is not what I have accomplished then I have failed to communicate.
 * —Trappist the monk (talk) 23:21, 16 March 2022 (UTC)
 * is inaccurate. If that is not what I have accomplished then I have failed to communicate.
 * —Trappist the monk (talk) 23:21, 16 March 2022 (UTC)

Should argument to via parameter allowed to have a url?
Given the January update to Module:Citation/CS1, I see that urls supplied to via now generate an error. However, common usage (e.g., FEIS and Calflora before I fixed them) and the documentation seem to imply that via should contain urls. In fact, I'm not sure how you can have non-url arguments?

Can we include via in the list of url-enabled parameters? — hike395 (talk) 21:39, 19 March 2022 (UTC)
 * Umm, common usage is not so common, at least according to these searches:
 * ~700 articles with http...
 * >219000 articles with &lt;anything> (search times out)
 * I don't read anything in the documentation that [implies] that via should contain urls.
 * —Trappist the monk (talk) 22:27, 19 March 2022 (UTC)
 * Via should not contain urls, no. &#32; Headbomb {t · c · p · b} 00:02, 20 March 2022 (UTC)
 * Respectfully,, that is not the right search query to use. If you look for where editors have used something url-like, you get
 * >16,000 articles.
 * A very common pattern is to refer to websites by name: if we look through your second query, you'll see usage like Yahoo News or Google Books (which are also suggested by the documentation). The main example in the documentation is Dictionary.com, which is a wikilink to partial url (a name of a website).
 * If via is being filled with partial URLs or names of external sites, why should links to external sites be excluded? Alternatively, if you want to maintain purity between name parameters and url parameters, I would suggest that we would need a via-url parameter. I think allowing URLs in via is simpler, but I'm open to either. — hike395 (talk) 00:29, 20 March 2022 (UTC)
 * The parameter is commonly used when there is an existing url in order to clarify the url's provenance (if not at the publishing entity) or the content's format (if different from the published-as-cited version). It may also be used to add clarity to a non-CS1-defined identifier. And it may wikilinked. Providing another url is clutter without obvious benefit. 68.173.76.118 (talk) 00:39, 20 March 2022 (UTC)
 * Personally, I also use the parameter for clarity in citable self-published works whether url exists or not. The publishing provider is inserted and optionally wikilinked. But even in these cases, I don't think a via-based url adds anything useful to the citation. 68.173.76.118 (talk) 01:10, 20 March 2022 (UTC)
 * If you care about the provenance of a citation, and it's not obvious from the url or publisher string, then it would seem providing an external link to the delivery site would be helpful, yes? That way readers can assess the delivery mechanism. Also, I'm not sure why a wikilink in a via argument would be acceptable while an external link would be clutter? — hike395 (talk) 01:23, 20 March 2022 (UTC)
 * Not the provenance of the citation, the provenance of the url. A reader who links to the url may be understandably mystified to see the work published by somebody other than the cited publisher. The presence of via indicates this is not accidental or erroneous. It is to be presumed that there are no copyvio issues and that the url publisher is at least as trustworthy as the cited publisher. Their treatment should be the same. Izno, below, elaborates. 64.18.9.194 (talk) 02:31, 20 March 2022 (UTC)
 * via as a parameter is essentially publisher, in which we do not allow URLs. I see no reason to do so here either. As for your [hike395]'s modified query, no, Trappist's is more correct. Yours finds a whole bunch of domain names with TLDs, when what you want is the URL to show what you think is a certain use. This is more correct than his, but still only some ~1500 uses, so that is not persuasive either.
 * I also do not see where in the documentation you believe that the documentation promoted using a URL in this parameter. Izno (talk) 01:50, 20 March 2022 (UTC)
 * My original statement wasn't clear --- I meant to say "website" not a syntactically correct URL. But given that no one wants to implement this, I'll withdraw the suggestion, and leave via deleted from the various botany templates. — hike395 (talk) 05:01, 20 March 2022 (UTC)

[u.a.]
Previous discussions:
 * Help talk:Citation Style 1/Archive 2
 * Help talk:Citation Style 1/Archive 28

I just stumbled across two citations in Dorsolateral prefrontal cortex (references 1 & 25) that use '[u.a.]' in location. Apparently '[u.a.]' is an initialism of unter anderem ('among other things') or und andere ('and others').

Also, apparently, '[u.a.]' was at one time used by WorldCat so was copied from there to here by Citoid and/or its predecessors into location. At this writing, there are about 2700 articles that have '[u.a.]'.

What to do with '[u.a.]'? Continue to ignore? Should Module:Citation/CS1 add a maint cat when '[u.a.]' is found? If we do that, what is the recommended maintenance? I suspect that we should at least document '[u.a.]' somewhere so that editors can know what it means.

—Trappist the monk (talk) 16:38, 20 March 2022 (UTC)


 * Assuming this is used exclusively for publisher location, I think it should be considered unnecessary and removed. We are well past the age where a publisher subsidiary would differentiate from another subsidiary of the same publisher and language. The primary lication of the imprint in the country of publication is sufficient. If the initialism is used (erroneously) in place of et al., that should be fixed.
 * Fire up the RfC for the required tracking category. 69.203.140.37 (talk) 18:55, 20 March 2022 (UTC)

Suggestion for generic author name
Regarding CS1 errors, I've noticed some incidences of "web master"/"webmaster"/"web-master" (and the capitalised variants). It may be useful to also detect those in the author, etc., fields. --Xurizuri (talk) 04:09, 19 March 2022 (UTC) (please ping me if you respond)


 * @Xurizuri - I've got BattyBot removing those values from the author parameter. Thanks for the suggestion! GoingBatty (talk) 21:37, 21 March 2022 (UTC)

SSRN limit increase
There exists a citation with a valid SSRN above the current configured limit of 4000000. See Paleobiota_of_the_Posidonia_Shale citation 89, or see here for a pasted version.

Thanks. Aidan9382 (talk) 20:59, 23 March 2022 (UTC)

Wikidata QID parameter
I have added the parameter q to the sandbox. WikiCite has been adding citations and linked bibliographic data to Wikidata, and linking to them would be useful. Can and should  be used? Can and should input be normalized to an uppercase "Q"? int21h (talk · contribs · email) 05:47, 24 March 2022 (UTC) int21h (talk · contribs · email) 06:34, 24 March 2022 (UTC)
 * How does the Q identifier help a reader discover the citation's source? 71.247.146.98 (talk) 11:46, 24 March 2022 (UTC)
 * There are several efforts to use QIDs and Wikidata, probably mostly related to WikiCite. Using as an example, a citation graph has been added using  to aid discovery, and projects like d:Wikidata:Scholia (WP) can visually explore the same QID e.g. Q24316383. Scholia might even be a better link than Wikidata itself. (Scholia was also recently Wikidata Data Reuse Days 2022.) There are also other tools such as Reasonator. int21h (talk · contribs · email) 13:58, 24 March 2022 (UTC)
 * I like building citation graphs as much as the next scientist, but how does that help the reader of an encyclopedia article find the reference they've been told to read? XOR&#39;easter (talk) 17:11, 24 March 2022 (UTC)
 * QIDs can provide literally thousands of data types (and thousands of data) to a reader to help find the reference they've been told to read, that neither this module nor any of its current external identifiers can provide. Wikidata further links to external id systems that this module and its current external ids do not. A limited selection is at d:Wikidata:List of properties/work. int21h (talk · contribs · email) 20:08, 24 March 2022 (UTC) int21h (talk · contribs · email) 21:20, 24 March 2022 (UTC)
 * As was pointed out in this section, there have been several discussions about Wikidata. Serious concerns have been raised about the provenance and reliability of Wikidata data, and the potential for circular references or/and self-references. That would be enough to disqualify its use in the verifiability process. Citations are not supposed to be pondered over or dwelled upon; if readers do so, then the citation is probably not exactly right. The point is to provide readers with the minimum of the most relevant metadata in order to easily and quickly verify the text and keep on reading (hopefully with renewed interest following the verification). Not to inundate them (or citation editors) with a bunch of data whose relevance and applicability is rare or uncertain. I don't see the usefulness of the Q identifier in this context. 71.247.146.98 (talk) 12:11, 25 March 2022 (UTC)
 * Where is the discussion that establishes a consensus for this change? I seem to recall that some sort of wikidata parameter has been discussed here before but no consensus to implement has ever been achieved.  Here are some discussions that did not achieve consensus:
 * —Trappist the monk (talk) 13:53, 24 March 2022 (UTC)
 * And thus the consensus building continues. I believe my comment above addresses how QIDs aid discoverability and exploration. What other issues came up? int21h (talk · contribs · email) 14:02, 24 March 2022 (UTC)
 * That is disingenuous. You have already been pointed to other issues with your program of converting references to QIDs very recently, in a discussion that you participated in at —David Eppstein (talk) 16:51, 24 March 2022 (UTC)
 * The discussion at WT:CITE is about an altogether different template and a different policy issue than brought up in the discussions above, AFAICT. WP:CITEVAR and discoverability are different issues. Did WP:CITEVAR come up in this module re QIDs? int21h (talk · contribs · email) 16:56, 24 March 2022 (UTC)
 * Please provide a summary of the discussion as of 25 March 2022 00:20 UTC. int21h (talk · contribs · email) 02:46, 25 March 2022 (UTC)
 * Other than yourself, it does not appear that any participants support inclusion of the proposed parameter. Nikkimaria (talk) 12:15, 25 March 2022 (UTC)
 * And thus the consensus building continues. I believe my comment above addresses how QIDs aid discoverability and exploration. What other issues came up? int21h (talk · contribs · email) 14:02, 24 March 2022 (UTC)
 * That is disingenuous. You have already been pointed to other issues with your program of converting references to QIDs very recently, in a discussion that you participated in at —David Eppstein (talk) 16:51, 24 March 2022 (UTC)
 * The discussion at WT:CITE is about an altogether different template and a different policy issue than brought up in the discussions above, AFAICT. WP:CITEVAR and discoverability are different issues. Did WP:CITEVAR come up in this module re QIDs? int21h (talk · contribs · email) 16:56, 24 March 2022 (UTC)
 * Please provide a summary of the discussion as of 25 March 2022 00:20 UTC. int21h (talk · contribs · email) 02:46, 25 March 2022 (UTC)
 * Other than yourself, it does not appear that any participants support inclusion of the proposed parameter. Nikkimaria (talk) 12:15, 25 March 2022 (UTC)

unfit url maintenance message
I keep seeing more and more of these green maintenance messages and I have been making an effort to fix them as I see them. But there is one type that cannot be fixed and that is url-status=usurped because I can't make the URL stop being usurped, e.g. this is the archive and this is the the original URL which surely meets the definition of "usurped" consisting of both spam and porn. So what do I do to make the green maintenance message go away? Otherwise I or others are going to keep seeing and try and fix it, again and again and again. I think a maintenance message has to be for things that can be fixed and where they can't be fixed, a way needs to exist to suppress the maintenance message to avoid the rework by others. The other thing is that such citations have some extra text appearing at the end of citation which isn't in the citation, which I would like to suppress. In the case of the URL above, it is "OurToowong page for BBC". The citation title is "BBC" and the citation website is "Our Toowong" so what is purpose of this extra text? I see it on the "Archived copy" title citations, where it makes sense (as it is usually a good guess for what the title should be), but I don't see its relevance when there is a genuine title in the citation. I note it appears to me when logged-out so the readers are seeing it too.Kerry (talk) 00:09, 6 March 2022 (UTC)
 * This is about Brisbane Boys' College and Toowong? Always good to say where you are seeing something so we don't have to guess...  On both of those pages the text 'OurToowong page for BBC' has been appended after the cs1|2 template's closing   and before the reference's closing  tag.  That text is not something the the template adds.  For Brisbane Boys' College, the 'OurToowong page for BBC' text was added at this edit.  Later, you converted the simple external link reference to use  at this edit.  Today, you added a reference and the 'OurToowong page for BBC' text to Twoong at this edit.
 * I think that you are the first to complain about lingering maintenance messaging.
 * —Trappist the monk (talk) 01:10, 6 March 2022 (UTC)
 * —Trappist the monk (talk) 01:10, 6 March 2022 (UTC)

"green maintenance message" .. I don't see a green message for cites with usurped. They don't really need fixing. Also in this edit you added deviated but the source URL is "404" - ie. dead, not deviated. -- Green  C  02:03, 6 March 2022 (UTC)
 * This example template has a green maintenance message:
 * User:GreenC/common.css does not have the css to enable the message display (I did not look at your various skin css pages). Without that css the maintenance messages will be hidden:
 * CS1 maint: unfit URL (link)
 * The classes  and   are defined in Module:Citation/CS1/styles.css.  To show maintenance messaging, see.
 * —Trappist the monk (talk) 12:02, 6 March 2022 (UTC)
 * I see. If by default the green messages are hidden from users, that's good because usurped URLs are not in need of maintenance and we shouldn't encourage users to do anything about it (by default). There might be an argument the cite or URL should be deleted entirely, but there are counter-arguments such as usurped URLs over time revert to 404s once spammers stop funding the domain name. My only thought is maybe we shouldn't green message usurped cites, however it's not a big deal either way. -- Green  C  14:42, 6 March 2022 (UTC)
 * —Trappist the monk (talk) 12:02, 6 March 2022 (UTC)
 * I see. If by default the green messages are hidden from users, that's good because usurped URLs are not in need of maintenance and we shouldn't encourage users to do anything about it (by default). There might be an argument the cite or URL should be deleted entirely, but there are counter-arguments such as usurped URLs over time revert to 404s once spammers stop funding the domain name. My only thought is maybe we shouldn't green message usurped cites, however it's not a big deal either way. -- Green  C  14:42, 6 March 2022 (UTC)

GreenC, Kerry, Trappist the monk: I don't mind as much the green text next to citations with unfit URL's, which doesn't even appear for users that don't enable it at their common CSS page. What I believe is more of a problem is the warning at the top of the page whenever an edit is previewed for pages containing an unfit URL (regardless of whether maintenance messages are enabled on a user's common page). I don't see any benefit to the warning which I think will just distract / take up the time of a lot of editors. I initially brought this issue up [https://en.wikipedia.org/w/index.php?title=Wikipedia:Help_desk&oldid=1080051664#Is_there_an_easy_way_to_view_maintenance_messages_for_an_article? here at Wikipedia:Help desk] because I couldn't figure out how to get the maintenances messages to display. The conversation then turned to why the warning is necessary in the first place, upon which Trappist directed me here. So I'm just throwing my 2 cents in on the issue.--Jamesy0627144 (talk) 02:08, 30 March 2022 (UTC)

Generic title
Hello, can "Please login" at the start of a title be added to the generic title list. Currently about 60 occurrences. Thanks, Keith D (talk) 14:59, 30 March 2022 (UTC)

New CS1 citation template
I'm very pleased to announce the newest addition to the CS1 family, Cite bathroom graffiti. &#123;{u&#124; Sdkb  }&#125;  talk 08:56, 1 April 2022 (UTC)

Generic author name "By"
|%2F&title=Special:Search&profile=advanced&fulltext=1&ns0=1&ns4=1&ns10=1&ns12=1 A value of "By" in last is present in 200+ articles at this time. Marking author/editor names starting with "By " or containing only "By" (but not detecting "Byrd" etc.) might be helpful. I don't think that "By" is a valid surname, but it is a hypothetical edge case. – Jonesey95 (talk) 17:02, 1 April 2022 (UTC)
 * John By would likely disagree. As would Oddbjørn By, Sverre By, and Eli Gunhild By. —David Eppstein (talk) 21:17, 1 April 2022 (UTC)
 * That person does not appear to be cited anywhere on Wikipedia. I did find Sverre By and Ulrika By cited in about five articles out of about 200. I will let consensus decide whether that ratio is worth implementing error detection at the cost of having to use () notation around two authors' last names. – Jonesey95 (talk) 21:41, 1 April 2022 (UTC)

Validity of URLs in inset parameter of cite map
A question that's come up recently and is now a point of conflict in some articles. The  parameter in cite map throws up an error if a URL is placed there, which is a common practice for maps that are divided between multiple pages/links (such as those issued by US state governments). Given that insets are analogous to chapters in books, I think we'd benefit from having an  parameter to hold links for insets that do require a separate link for verifiability. I posted about this a few months ago but it didn't gain any traction, so I hope there'd be some feedback this time around.  Sounder Bruce  21:36, 2 April 2022 (UTC)
 * If we can link a page number for a book, we should be able to link an inset on a map.  Imzadi 1979  →   22:12, 2 April 2022 (UTC)
 * A change like this will resolve the error with no change in visible output (other than removal of the red error message). DrKay (talk) 07:21, 3 April 2022 (UTC)


 * Using |at= :
 * Using |inset= :
 * Given that insets are analogous to chapters in books,... Although this may not be strictly true in the sense that insets are more akin to sections rather than chapters, there is (by design) information in inset maps that does not appear in the base map. This is enough to warrant not only a parameter but also a convenience-link option. In proper software design, the cite map inset would be an alias of (hypothetical base-template) section, inset-url would be an alias of section-url, and so on. In any case the OP's request seems justified, at least for print maps, although perhaps it should be considered for digital/online maps as well. Links:.
 * 65.88.88.201 (talk) 15:14, 3 April 2022 (UTC)

eISBN again
Should it be reconsidered as a dedicated identifier? Some data providers require the field from publishers eg JSTOR Metadata Requirements: "Electronic ISBN. We require an eISBN for each book. We can not use the print ISBN as the electronic ISBN." Which means that a citation with the addition of eISBN could expedite discovery. 65.88.88.93 (talk) 19:37, 4 April 2022 (UTC)

In case it was not obvious, this concerns works published in digital media (including eBooks), regardless of whether they are also published in print. 65.88.88.93 (talk) 19:55, 4 April 2022 (UTC)

Adding internet image links
Hello. When I expanded the section of an article, I used the cite magazine template to cite an interview and I added direct links to scans of those pages, so the content can be verified. To be honest, however, I'm not entirely sure about this practice and a potential copyright infringement. The magazine is defunct, so it would be very hard to acquire it physically. I suppose that I could simply use the cite magazine template without adding links, but I would rather prefer that the content can be easily checked. I would like to know some inputs about this. - Xexerss (talk) 22:40, 4 April 2022 (UTC)
 * There is guidance at WP:COPYLINK. – Jonesey95 (talk) 22:52, 4 April 2022 (UTC)

Params transcript & transcript-url
I don't know exactly when or why they disappeared, and whatever nonsense preceded the disappearance is not important. They are not trivial, and are needed now. The hacks and work-arounds have run their course, the parameters themselves should reappear. Citations that traditionally and currently need these parameters are usually input by editors with the following templates: 65.88.88.57 (talk) 18:07, 5 April 2022 (UTC)
 * cite AV media
 * cite conference
 * cite episode
 * cite interview
 * cite podcast
 * cite speech
 * I think this September 2020 discussion describes the most recent change that is relevant to this question. Here are examples of those six templates (plus cite serial, whose documentation currently contains transcript), showing that at this writing, AV media and episode work fine.


 * I think the next step is to provide examples of actual citations of sources where transcript would be used in real articles. – Jonesey95 (talk) 18:38, 5 April 2022 (UTC)
 * All six templates are used to cite sources originally in non-text media. Is an example really needed to demonstrate the usefulness (re:verification/accessibility etc.) of providing the content in text form in a unique, format-specific parameter? One may very well have a citation of a radio/tv interview with a URL to the stream and a URL to the transcript. Or a speech citation that provides links to both the audio and/or video and the text. 65.88.88.57 (talk) 18:56, 5 April 2022 (UTC)
 * Templates are generally constructed to meet actual, not hypothetical needs. It should be easy to demonstrate that need, for the record. I believe that you or someone else can do it here with little effort, with the exception that cite conference is for published proceedings, not for audio/video sources. – Jonesey95 (talk) 20:39, 5 April 2022 (UTC)
 * The enhanced verifiability a transcript provides is not hypothetical, but actual and actionable. Citations may include multiple content-locators such as doi/pmc/jstor etc. This case is similar. And as things stand, I cannot provide a citation example without generating an error. By the way, conference proceedings may also be published on DVD video. 65.88.88.57 (talk) 21:27, 5 April 2022 (UTC)
 * cite news is another template that could use these parameters. Both television/radio and web video/radio news shows may offer transcripts. 4.30.91.142 (talk) 20:23, 6 April 2022 (UTC)

Protected edit request on 7 April 2022
Remove "unique" and add: They have also been found among a Northern Jê speaking people, the Mẽbêngôkre in Central Brazil
 * I suspect this is in the wrong place. Where might we "Remove 'unique' and ..."? It would be best if you could provide a link like Kinship terminology. Thank you, SchreiberBike &#124; ⌨ 19:42, 7 April 2022 (UTC)
 * Red information icon with gradient background.svg Not done: this is the talk page for discussing improvements to the page Help:Citation Style 1. Please make your request at the talk page for the article concerned. Izno (talk) 20:40, 7 April 2022 (UTC)
 * For the record, the above request was nominally about Cite book, but it was misplaced, as indicated by . – Jonesey95 (talk) 21:36, 7 April 2022 (UTC)

Category:CS1 maint: ref=harv
Before, we had this category to track ref=harv usage. Apparently now it's been merged to Category:CS1 errors: invalid parameter value


 * https://en.wikipedia.org/w/index.php?title=Eulogy_of_King_Prasat_Thong&diff=prev&oldid=1081637277
 * https://en.wikipedia.org/w/index.php?title=Daniel_Trotter&diff=prev&oldid=1081637337

However, the prior version of these diffs weren't caught by that category. Is suspect it's because harv was case sensitive, but really it should be case insensitive. &#32; Headbomb {t · c · p · b} 17:41, 8 April 2022 (UTC)
 * It is probably a good idea to do a NOT search, such as for  values that do not start with   or   etc. And fire up the RfC for the related tracking cat. 74.64.30.159 (talk) 19:48, 8 April 2022 (UTC)

Special-issue title of an ongoing journal
Sometimes journals release a "special issue" on a certain theme, either one issue on their regular release-cycle or an off-cycle release. It sometimes invites guest-editors to participate, invites specific authors to contribute to that special issue, or else selects among the regular submissions at the time to include (rather than for example in order of submission or editorial acceptance). These special issues have a subtitle. Should this subtitle be included in the citation? If so, where does it go in cite journal (series, journal, somewhere else)? Does it matter if a journal very often has these sorts of issues (a subtitle of nearly every issue) or only on a rare occasion?

As an example, 10.1016/j.chroma.2004.08.072 is in the special theme issue "Food Science" of Journal of Chromatography A. DMacks (talk) 02:37, 11 April 2022 (UTC)
 * The theme of special issues is bibliographically irrelevant, so no it should not be included. &#32; Headbomb {t · c · p · b} 02:47, 11 April 2022 (UTC)
 * Makes sense to me. Thanks! DMacks (talk) 05:54, 11 April 2022 (UTC)
 * If you must include it, for an individual article, it might be possible to use the department parameter for it. That has the advantage of not being passed on to the machine-readable side of the metadata, and therefore not messing it up. But I tend to agree with Headbomb in not mentioning it at all, for most citations. (If it's in your own cv, then maybe it's relevant information, but not for most other uses of citations, and we don't host cvs here.) If you're including the whole issue among the selected publications of an academic, then you could just use the title parameter. —David Eppstein (talk) 06:47, 11 April 2022 (UTC)
 * I would not call the special issue a "subtitle". It is an issue . If you are so inclined, you may quote it with Special issue name. 172.254.222.178 (talk) 11:36, 11 April 2022 (UTC)
 * Sorry, this was not as clear as I had hoped. A more obvious (I hope) example:
 * 71.247.146.98 (talk) 12:52, 11 April 2022 (UTC)
 * Don't abuse parameters like that. These are neither departments, nor quotes. &#32; Headbomb {t · c · p · b} 15:05, 11 April 2022 (UTC)
 * Actually, anything cited from the source as specific to the particular edition/issue, such as an issue name or an article title is ... quoted, and should be in quotation marks. Unlike items like page numbers that are typesetting artifacts, sections such as Foreword that are layout artifacts or items like volume numbers that are serial artifacts. Special issues are usually prepared by editorial teams specific to the purpose. "Editorial team" and "Department" are synonymous in practice. 50.75.226.250 (talk) 15:27, 11 April 2022 (UTC)
 * For the curious: Chicago 16 §14.178 recommends citations like Miwako Tezuka, "Jikken Kōbō and Takiguchi Shūzō: The New Deal Collectivism of 1950s Japan," in "Collectivism in Twentieth-Century Japanese Art," ed. Reiko Tomii and Midori Yoshimoto, special issue, Positions: Asia Critique 21, no. 2 (Spring 2013): 351–81, https://doi.org/10.1215/10679847-2018283., including the title of the special issue (as well as editors), but APA 7 §10.1.12 reads For an article within a special section or special issue, follow the format for a journal article [...], in which case the title of the special section or issue does not appear in the references.
 * I personally think when there's a DOI, that extra information like title of a special issue doesn't do more to help a reader locate a source. But for older, print-only sources, it can be useful to provide more information. Sometimes, libraries will have separate entries for special issues. There is a separate for Food Science, eds. Careri & Robards -- granted this is only used by two libraries in Germany so I don't think this is particularly useful, but in other cases knowing the title of the special issue will be useful in helping track down copies. It might be worth seeing how others have treated the particular article -- I just checked Google Scholar and the first three papers I found citing your source don't include the special issue title or editors, which I think is a good sign you don't need to include that in your citation either.
 * Also, as a note, in other cases, the line between book series and journals is thin. This isn't true in all cases, but sometimes what might look like a journal is more of a book series, in which case cite book is almost perfect (if only it had a issue -- actual monographs in actual book series have had issues and volumes and there's no good way to include that information in a cite book). Umimmak (talk) Umimmak (talk) 20:25, 11 April 2022 (UTC)
 * A book series (of multiple titles/volumes) and a journal (of multiple issues) are two completely different things, and AFAIK, no metadata provider in any format, has ever bundled the two. There is an overall classification of "serials" (now also called "continuing resources") that includes journals and single-title books published (serialized) a practice that is now rare in print, although sometimes found in online publishing. DOIs are content (in-source) identifiers that the majority of readers is likely ignorant of, although source identifiers like ISBNs may be marginally better known to the general public. But the OP was about an entire issue. There seems to be a consensus that the special name info is helpful, though not strictly necessary, in Wikipedia citations. 65.88.88.62 (talk) 21:27, 11 April 2022 (UTC)
 * Also, as a note, in other cases, the line between book series and journals is thin. This isn't true in all cases, but sometimes what might look like a journal is more of a book series, in which case cite book is almost perfect (if only it had a issue -- actual monographs in actual book series have had issues and volumes and there's no good way to include that information in a cite book). Umimmak (talk) Umimmak (talk) 20:25, 11 April 2022 (UTC)
 * A book series (of multiple titles/volumes) and a journal (of multiple issues) are two completely different things, and AFAIK, no metadata provider in any format, has ever bundled the two. There is an overall classification of "serials" (now also called "continuing resources") that includes journals and single-title books published (serialized) a practice that is now rare in print, although sometimes found in online publishing. DOIs are content (in-source) identifiers that the majority of readers is likely ignorant of, although source identifiers like ISBNs may be marginally better known to the general public. But the OP was about an entire issue. There seems to be a consensus that the special name info is helpful, though not strictly necessary, in Wikipedia citations. 65.88.88.62 (talk) 21:27, 11 April 2022 (UTC)

ONIX
ONIX is now the international defacto metadata transmission and formatting standard for descriptive bibliographic information. It is administered by EDItEUR. The current major version is 3, with revision number 08. ONIX for Books is one of the standards, pertaining to non-serial items, although the definition of "serial" is complicated. ONIX bibliographic records are XML documents built with data vocabularies that are actively developed. The vocabularies contain "codelists". These are lists of codes that represent fields in the bibliographic record. For example, List 5 is labelled "Product identifier type" and contains codes of identifiers such as ISBN, DOI etc. along with descriptions. The codelists are frequently amended, and are published by EDItEUR as "Codelist Issues". The current issue is Issue #56, and contains over 250 codelists with literally thousands of codes (or fields). Structured citations are subsets of bibliographic records that may contain all or some of the relatively very few fields of the record that aid in the recorded item's speedy discovery and acquisition. I post this to aid further discussion here on the semantics and nomenclature employed by CS1 and other structured citations in Wikipedia, and to harmonise them as much as possible with the nomenclature of established global standards under the following constraint:

Links 50.75.226.250 (talk) 15:38, 12 April 2022 (UTC)
 * ONIX Resources
 * ONIX codelists
 * A few ONIX implementers
 * A few ONIX implementers


 * So... what do you actually want? Pick a specific change you might like to see. Izno (talk) 17:23, 12 April 2022 (UTC)
 * I think the rationale for posting was explained? It is to be thought of as an/one authoritative reference for the various discussions on this page. Assume a new parameter is discussed. Obviously it would be helpful if a corresponding field existed in the major bibliographic classification standards. If it does exist, then the parameter usage, label, and description can take this info into account. Hopefully it is going to result in fewer/shorter discussions, and better align with the standards (within the Wikipedia constraints). 68.174.121.16 (talk) 19:21, 12 April 2022 (UTC)
 * I don't want anything to do with the terms and conditions. Jc3s5h (talk) 21:08, 12 April 2022 (UTC)
 * ? Care to expand? Although the scheme is copyrighted, there is no licensing involved. And this was proposed as a reference, not in any other capacity. 65.254.10.26 (talk) 00:16, 13 April 2022 (UTC)
 * I meant there is no payment involved in the license. I don't see any other onerous conditions but I'm not a legal expert. 65.254.10.26 (talk) 00:30, 13 April 2022 (UTC)

Category:CS1: long volume value has been nominated for deletion
Category:CS1: long volume value has been nominated for deletion. A discussion is taking place to decide whether this proposal complies with the categorization guidelines. If you would like to participate in the discussion, you are invited to add your comments at the category's entry on the categories for discussion page. – Jonesey95 (talk) 17:59, 12 April 2022 (UTC)
 * Please note: I did not nominate this category for deletion. This is a courtesy notice. – Jonesey95 (talk) 17:59, 12 April 2022 (UTC)
 * Bah! Fire up the RfC. Once CS1 development became contingent on largely uninformed opinions, this was to be expected. Also, counterpropose tracking cat "Category: RfCs related to CS1/CS2 tracking categories". To include the RfC for the creation of the counterproposed cat, ofcourse. 68.174.121.16 (talk) 18:28, 12 April 2022 (UTC)

Increase PMID limit to 40000000


This shouldn't throw an error. &#32; Headbomb {t · c · p · b} 01:29, 14 April 2022 (UTC)

Newspaper sections
Greetings! I added this citation today to West Springfield, Massachusetts:

I'm not sure why it's complaining "section= ignored"; paper newspapers have sections with letters or names or numbers which need to be included to distinguish the page numbers. Is there some other way to do this properly, or is there a template bug? Thanks! -- Beland (talk) 03:06, 12 April 2022 (UTC)


 * For what you’re using "section" for, note If hyphenated, use hyphen to indicate this is intentional (e.g. 312), otherwise several editors and semi-automated tools will assume this was a misuse of the parameter to indicate a page range and will convert 3-12 to 3–12. so I’d use 130. You I guess also could use §1, p. 30 or something like that. Umimmak (talk) 03:11, 12 April 2022 (UTC)
 * And, if you would like to have the name of the section, use department. This is one of its supported uses. Izno (talk) 17:22, 12 April 2022 (UTC)
 * Unfortunately, department comes before the same of the newspaper, so:
 * gives you:
 * — Preceding unsigned comment added by Hawkeye7 (talk • contribs)
 * The presumption when I said what I said was an actual word like "Living", not the letter denoting one part of the page number. --Izno (talk) 23:43, 13 April 2022 (UTC)
 * As suggested above, use at, but do not use the section symbol, it signifies a page section. The "section" here refers to a section of the work, which in newspapers may be bound separately. Also, the NYT used to paginate according to section+page, as in (page) A30. If this the case, you can just enter the page, as the section ("A") is included. 74.64.150.19 (talk) 12:12, 12 April 2022 (UTC)
 * The presumption when I said what I said was an actual word like "Living", not the letter denoting one part of the page number. --Izno (talk) 23:43, 13 April 2022 (UTC)
 * As suggested above, use at, but do not use the section symbol, it signifies a page section. The "section" here refers to a section of the work, which in newspapers may be bound separately. Also, the NYT used to paginate according to section+page, as in (page) A30. If this the case, you can just enter the page, as the section ("A") is included. 74.64.150.19 (talk) 12:12, 12 April 2022 (UTC)

Where should I put a translation link
For "cite web" or "cite interview" templates: Let's say the source is in a foreign language, but an English translation is available elsewhere. Is there a parameter that I can use to add the English translation? Thanks.--TerryAlex (talk) 01:15, 14 April 2022 (UTC)
 * One way is to use the English translation as the source. Add originally published in year in language. Also use translator-last etc. if you know the translator. 64.18.9.194 (talk) 03:21, 14 April 2022 (UTC)
 * Cite the version that you are citing, per WP:SAYWHEREYOUGOTIT. You can add a normal URL link after the citation template, but before the tag. – Jonesey95 (talk) 04:45, 14 April 2022 (UTC)
 * If both the original and the translation are available, and the translation is trusted, the translation should be used, with the link in url. So that English Wikipedia readers are able to verify the related text. 64.18.9.194 (talk) 11:26, 14 April 2022 (UTC)

et al and punctuation
Because I found an article that had explicit 'et al' text in a name holding parameter that was not emitting an error message, I have tweaked the sandbox so that semicolons are detected by the etal pattern detector:

The pattern already detects,  ,  , and. This change just adds.

—Trappist the monk (talk) 18:42, 15 April 2022 (UTC)

Script warning shown at top of preview, without error message or category
When I use the source editor on Torre de Don Miguel, I am shown a script warning telling me that one or more cite journal templates have errors, but I see no error message in the previewed article, and I see no CS1 category in the rendered article. I am pretty sure that I have all of the error messages enabled. – Jonesey95 (talk) 15:06, 16 April 2022 (UTC)
 * In the infobox are and .  That template calls  which in turn calls  which renders this:
 * : More than one of  and   specified (help)
 * But, only tests that  returns something:  .   template is created but not rendered.  The creation adds the error message to cs1|2's internal list of error messages even though the citation and category are not rendered.  cs1|2 then, dutifully shows the preview warning because it cannot know that the citation isn't rendered.  You can mimic this by adding a parameter that is unknown to the infobox with a cs1|2 template as the parameter's value:
 * unknown
 * now preview; no rendered citation, no url-required error message, but there is a preview message.
 * wrote a citation with both at and pages.
 * —Trappist the monk (talk) 15:52, 16 April 2022 (UTC)
 * This search finds about 7400 articles that may have this problem.
 * —Trappist the monk (talk) 16:06, 16 April 2022 (UTC)
 * Presumably, a fix can be applied at that changes:
 * to:
 * After all, the presence of a reference is no guarantee of the presence of the property ...
 * —Trappist the monk (talk) 16:20, 16 April 2022 (UTC)
 * Perhaps it is better not to touch this whole situation with a 10-foot pole, but just would observe that cite report not is likely the proper record/resource type here. 68.160.224.18 (talk) 17:43, 16 April 2022 (UTC)
 * to:
 * After all, the presence of a reference is no guarantee of the presence of the property ...
 * —Trappist the monk (talk) 16:20, 16 April 2022 (UTC)
 * Perhaps it is better not to touch this whole situation with a 10-foot pole, but just would observe that cite report not is likely the proper record/resource type here. 68.160.224.18 (talk) 17:43, 16 April 2022 (UTC)
 * Perhaps it is better not to touch this whole situation with a 10-foot pole, but just would observe that cite report not is likely the proper record/resource type here. 68.160.224.18 (talk) 17:43, 16 April 2022 (UTC)

Cite Web Wrapper at TfD - Cite CinemaScore
Hey, Template: Cite CinemaScore is a template that wraps Cite Web, and has been nominated for deletion where you may wish to comment Templates for discussion/Log/2022 April 17 Indagate (talk) 14:55, 18 April 2022 (UTC)

Updating the live version
Is there a timeline for when the CS1 modules get updated? If there isn't one, could we update them with the sandbox changes? Gonnym (talk) 09:58, 21 April 2022 (UTC)

|location= without |publisher=
More often these days I have noticed that templates created through visual editor have location but omit publisher. For example, this template, created at this edit, at Bipedidae:

It isn't just visual editor, using WP:RefToolbar and autofilling from the ISBN will also create a template with location but without publisher.
 * It was my understanding that both ve and reftoolbar both use citoid but if that is true, it is interesting that the results are astonishingly dissimilar.
 * It was my understanding that both ve and reftoolbar both use citoid but if that is true, it is interesting that the results are astonishingly dissimilar.
 * It was my understanding that both ve and reftoolbar both use citoid but if that is true, it is interesting that the results are astonishingly dissimilar.

I begin to wonder if cs1|2 should emit an error message for templates and for  templates without a work alias when location has a value but publisher is omitted or empty.

Opinions?

—Trappist the monk (talk) 14:15, 23 April 2022 (UTC)


 * I would tend to agree; I'm not coming up with a situation where it's valid to have location and not publisher. I suppose there are questions around the best way to represent self-published books, but even then publisher should say something (self-published, privately printed, whatever). Mackensen (talk) 14:22, 23 April 2022 (UTC)
 * Agree with the error message for any template and not just . But only after csdoc regarding § Publisher is written properly. The documentation includes a rarely-useful (for discovery purposes) author-related parameter (created/written at) under the publisher section, and erroneously calls it an alias of location/publication-place, both of which refer to the publisher/imprint location. The latter parameters are very often included in bibliographic records and although their usefulness in discovery is marginal and their presence not necessary, they make no sense without publisher. 50.75.226.250 (talk) 14:38, 23 April 2022 (UTC)
 * The documentation has long discouraged the use of publisher in some CS1 templates like cite news (Not normally used for periodicals. Omit where the publisher's name is substantially the same as the name of the work ...), so applying this requirement to templates other than cite book is probably not a good idea. Even for books, I suspect that this requirement would turn out to be overly fussy, with many situations in which publisher information is unavailable. We don't want unfixable error messages. What would our recommendation be in that case? none would have to be accepted, but I think there would be complaints of the type "there was clearly a publisher of some kind, but the information doesn't exist, so 'none' is untruthful." – Jonesey95 (talk) 14:47, 23 April 2022 (UTC)
 * The proposed error message as I understand it, is to point out the error of including location without publisher. Unless Trappist means something else. Also, assuming the documentation re: place etc. is corrected, the module could be edited to make the location arg conditional on publisher. 50.75.226.250 (talk) 15:11, 23 April 2022 (UTC)
 * Yeah, and, without work aliases,  though, , , , , and  might also be checked.  For the nonce,  and .  I can imagine a sort of similar case for  which should not need and really shouldn't support, location.
 * It is relatively easy when using cirrus search to find something but not so easy to find the absence of something. I've asked at  to see if there is a way to discover the magnitude of the issue beforehand.  We can always start out with a maint cat and at a later date migrate to error messaging.
 * —Trappist the monk (talk) 15:59, 23 April 2022 (UTC)
 * Apparently it is not possible to do quarry searches of wikitext so that idea has fizzled.
 * —Trappist the monk (talk) 19:28, 23 April 2022 (UTC)
 * Re: self-published works. The majority of modern self-published works are rarely self-published. Professional services are contracted by authors (instead of the normal, other way-around) and they may be involved in both the physical/digital technical, distribution and marketing of the work. In these cases, the parameter publishing service is likely useful, as something similar would be included in cataloguing/classifiications of the work. 50.75.226.250 (talk) 14:46, 23 April 2022 (UTC)
 * This proposed use of via conflicts with our documentation. – Jonesey95 (talk) 14:49, 23 April 2022 (UTC)
 * ? Where is the conflict? The content-deliverer (the publishing service) is different from the publisher (the author, in self-published works). Also, in the real world, the names of such services are recorded for cataloguing purposes. 50.75.226.250 (talk) 15:00, 23 April 2022 (UTC)
 * I would not mis-use via that way. We should still list the vanity press as the publisher.  Imzadi 1979  →   16:54, 23 April 2022 (UTC)
 * It doesn't work that way in an unreliable platform like Wikipedia. Wikitext statements and their supporting references have reliability, notability and neutrality requirements. As self-published sources are not prohibited in Wikipedia, such sources must explicitly identify as such. Omission slants reader evaluation of the offered information. I don't see any mis-use of via here. Not only is the formulation factual and within the use-cases, it provides important information about the nature of the source to the reader. 64.18.9.197 (talk) 23:56, 23 April 2022 (UTC)
 * Citations are not only for modern books. I have seen plenty of 17th and 18th-century book citations where the city of publication is known but there is no specific publisher listed. This continued push to make the citation templates as inflexible as possible is unhelpful. —David Eppstein (talk) 16:05, 23 April 2022 (UTC)
 * I'm drawing a blank on which established citation style said so, but one (or more) say to list location on old works such as that and omit the publisher, if known.  Imzadi 1979  →   16:54, 23 April 2022 (UTC)
 * I did not do an extensive search but, according to these websites:
 * APA uses '(n.p.)' when missing publisher or when missing both publisher and location
 * Harvard style uses '[no publisher]' and '[no place]'
 * Chicago style uses 'n.p.' (for both) or abbreviation of the Latin sine loco and sine nomine, 's.l.' and 's.n.'
 * No doubt, cs1|2 can adopt something similar if given, for example, none and/or none. This is not an insurmountable problem.
 * —Trappist the monk (talk) 19:28, 23 April 2022 (UTC)
 * It is not a problem at all. You are only making it a problem, by making it harder for anyone but a bot to properly format a correct citation, when a problem does not already exist. Instead just make a tracking category and check the results manually. It doesn't need to be an error. —David Eppstein (talk) 20:15, 23 April 2022 (UTC)
 * Yeah I had the same thoughts regarding old books, so I'm glad you made that point. Umimmak (talk) 03:42, 24 April 2022 (UTC)
 * However, sources must also be available as far as possible. How is a reader to verify text in an 18th-century book? If the work is notable, modern editions/reprints/fascimiles should be more generally available. Even assuming that the editor has access to such rare originals it is probably better to cite an available, trusted modern reprint. 65.88.88.201 (talk) 14:56, 24 April 2022 (UTC)
 * As long as the reference is verifiable, then just that it is old shouldn't preclude it from being used as a reference. For example there are plenty of old works on the Internet Archive, and in major libraries.Nigel Ish (talk) 15:14, 24 April 2022 (UTC)
 * But for a reference to be verified the source must first be discovered. Obviously facsimiles such as those provided by the Internet Archive's official archivists should probably be considered reliable reprints (sorry, I have found scan errors/missing sections in in-house IA scans as well). In contrast, third-party uploads to IA are questionable and should not be considered a priori reliable. I wonder how many libraries have original issues of 18th-century works... Isn't it just easier for everybody to use a readily available trusted reprint? Citations in Wikipedia are not there to support research on a subject, but to help someone quickly verify the text. 65.88.88.201 (talk) 15:37, 24 April 2022 (UTC)
 * Please reread WP:SOURCEACCESS - "Do not reject reliable sources just because they are difficult or costly to access. Some reliable sources are not easily accessible. For example, an online source may require payment, and a print-only source may be available only through libraries. Rare historical sources may even be available only in special museum collections and archives." - citation tools should not be used to overrule Policy, but should support it.Nigel Ish (talk) 18:09, 24 April 2022 (UTC)
 * In many cases old books can be found online at archive.org. That does not make them new publications with archive.org as the publisher. They are still the same old book they were; archive.org is at best a via parameter rather than a publisher. —David Eppstein (talk) 16:05, 24 April 2022 (UTC)
 * This is not disputed, assuming the content delivery source is trusted. There was also the valid point of older works with a location and not publisher. The proposed solution was to use a modern, accessible reprint, which should be available if the original was in any way notable. Unless of course the original edition was specifically referred to in text (eg in an article about the original). Nobody is rejecting valid sources. But if the source cannot be consulted by verifiers in any form, original or reprint, the text cannot be verified and WP:V is violated. 50.74.21.22 (talk) 18:40, 24 April 2022 (UTC)
 * Sorry to butt in here. Trappist, the specialized citation systems you refer to above follow the classification shorthand of the cataloguing/trade/bibliographic resources they were based on, which for the US was mainly (but not solely) the Library of Congress cataloguing system. Publishers had to provide bibliographic information to LoC as one of the requirements in order to be quickly assigned copyright, based on a Library-assigned identifier. That was before the British SBN system became a global standard. For the purposes of Wikipedia (general audience) it is better to not use such shorthand whenever possible, or at least to provide a parallel formulation more understandable to the reader. 65.88.88.201 (talk) 15:14, 24 April 2022 (UTC)
 * Um, specialized? APA and Chicago are two of the external style guides from which cs1|2 gets its style.  I was thinking that we could do something like none rendering like this:
 * Title. San Francisco: [no publisher]. – simulation
 * I was not suggesting that we adopt (n.p.) etc as something for editors to write. Editor confusion has been seen with n.d. and n.d. so it might be better if we deprecate   and   and replace those keywords with   so:
 * Title. [no date] – simulation
 * EB Green ([no date]). Title. – simulation
 * In, leaving publisher blank when location has a value might be automatically treated as if none were present plus an attendant category and maint/error message (non cat / no message when none and / or none explicitly stated).
 * —Trappist the monk (talk) 16:11, 24 April 2022 (UTC)
 * Yes, it is unfortunate that a citation style for general readership (a first) has to follow systems geared to an expert readership, with Chicago styles geared to either the humanities sector or the sciences sector, and APA having an even narrower focus. The entire basis is wrong, but it can be fixed to match the readership. Whether that happens is a different story. 50.74.21.22 (talk) 18:26, 24 April 2022 (UTC)
 * Title. [no date] – simulation
 * EB Green ([no date]). Title. – simulation
 * In, leaving publisher blank when location has a value might be automatically treated as if none were present plus an attendant category and maint/error message (non cat / no message when none and / or none explicitly stated).
 * —Trappist the monk (talk) 16:11, 24 April 2022 (UTC)
 * Yes, it is unfortunate that a citation style for general readership (a first) has to follow systems geared to an expert readership, with Chicago styles geared to either the humanities sector or the sciences sector, and APA having an even narrower focus. The entire basis is wrong, but it can be fixed to match the readership. Whether that happens is a different story. 50.74.21.22 (talk) 18:26, 24 April 2022 (UTC)
 * Yes, it is unfortunate that a citation style for general readership (a first) has to follow systems geared to an expert readership, with Chicago styles geared to either the humanities sector or the sciences sector, and APA having an even narrower focus. The entire basis is wrong, but it can be fixed to match the readership. Whether that happens is a different story. 50.74.21.22 (talk) 18:26, 24 April 2022 (UTC)

Can sfn pull data from a CS1 template?
I was curious: Is it possible for sfn to pull metadata from a parent citation like cite book? I.e., if there was a parameter at the parent citation level for the Google Books or Internet Archive identifier, could sfn, which holds a specific page number, generate a direct link? E.g., d4v3QgfhPKwC in the parent template, and 243 in sfn, would together output "https://books.google.com/books?id=d4v3QgfhPKwC&pg=PA243" in the short footnote. This would safe a lot of manual text in how page numbers are currently linked in sfn templates and make breakages easier to fix. czar 01:41, 24 April 2022 (UTC)


 * No, this is something MediaWiki does not support and any such attempts to "cross-talk" between templates even of the same kind have been removed in the past by the developers (i.e. are definitely not supported). At best with getContent you're talking something fragile as heck, and I don't even think that could do it today. Izno (talk) 04:08, 24 April 2022 (UTC)
 * Thank you. Out of curiosity and at risk of going off-topic, how is CS1 able to format its dates based on the existence of another date template elsewhere in the article, or is that an exception? czar  04:13, 24 April 2022 (UTC)
 * There is a Lua function called (roughly?) getContent, which gets the wikitext content of the page. Then it looks for the wikitext of interest. This is reasonably cheap because those are usually at the top of the article, and it's only done once a page in our case. It's probably something that won't be supported at some point in the indeterminate future, at least for same-page use, due directly to aforementioned issues, but on top of that the sfn version of the same would be much more expensive probably. Izno (talk) 04:34, 24 April 2022 (UTC)

Series volume vs. volume
I suggest the template needs a series-volume parameter. Many series are organized in volumes, and some editors are changing The OdysseyMasterworks, vol. 152 (-> ) to The OdysseyMasterworks152 (-> ), which is misleading. The problem becomes worse, and beyond the template's capability, if the work itself is published in volumes:. Our German colleagues have implemented that at de:Vorlage:Literatur with BandReihe [volume series]. -- Michael Bednarek (talk) 06:49, 30 April 2022 (UTC)
 * Not sure if I understand where the problem is. Why is the second rendering misleading? Isn't the series name "Masterworks"? It is pretty obvious that "volume" refers to the series volume. 74.64.150.19 (talk) 12:42, 30 April 2022 (UTC)
 * It seems I constructed the examples not clearly enough. If series-volume and volume are closer in value to each other, the problem is more obvious. -- Michael Bednarek (talk) 13:52, 30 April 2022 (UTC)

Semi-protected edit request on 30 April 2022
This is a blank edit request, please advise if amend is requested? Thanks, Indagate (talk) 13:11, 30 April 2022 (UTC)

Template parameter assistance
At Cite web, I am having trouble with the "|url-access=" parameter to say it requires a subscription at Jordan Poole citation 82.-TonyTheTiger (T / C / WP:FOUR / WP:CHICAGO / WP:WAWARD) 23:45, 3 May 2022 (UTC)
 * This looks moderately correct:
 * Using is probably a better choice.  Instead of Sports (an alias of chapter which is not supported in  and ) use Sports
 * Since what you wrote is moderately correct, what is the problem that you are having?
 * —Trappist the monk (talk) 00:07, 4 May 2022 (UTC)
 * When I mouseover citation 82 in the text, I don't see the red lock. Now that I am thinking to look down in the references, I see what you are saying. I guess my problem is that the mouseover does not present the same as the refence section at the bottom and I did not know that there was a difference. I guess it is working correctly. — Preceding unsigned comment added by TonyTheTiger (talk • contribs) 01:40, 4 May 2022 (UTC)
 * If you are using Tools/Navigation popups, yeah, you won't see the access icon. That tool, apparently, doesn't understand css.
 * —Trappist the monk (talk) 02:52, 4 May 2022 (UTC)
 * When I mouseover citation 82 in the text, I don't see the red lock. Now that I am thinking to look down in the references, I see what you are saying. I guess my problem is that the mouseover does not present the same as the refence section at the bottom and I did not know that there was a difference. I guess it is working correctly. — Preceding unsigned comment added by TonyTheTiger (talk • contribs) 01:40, 4 May 2022 (UTC)
 * If you are using Tools/Navigation popups, yeah, you won't see the access icon. That tool, apparently, doesn't understand css.
 * —Trappist the monk (talk) 02:52, 4 May 2022 (UTC)

Proposal: add or   parameter
Some web pages have a precise beginning date but are continuously updated. It makes no sense in these cases to use the date parameter, but it would make much more sense to implement a since or (alternatively) a since-date parameter. This would be displayed, using 15 January 2022 as an example, as “(since 15 January 2022)”. --Grufo (talk) 14:20, 30 April 2022 (UTC)
 * Wouldn't you just use access-date? The original date, if known, doesn't seem relevant. Indagate (talk) 14:25, 30 April 2022 (UTC)
 * If pages are constantly updated and the website is maintained properly, then it is likely that the update date will be published on the page. In this case, update-date. If an update date is not listed, the options are not good. You can use a copyright date if available on the page. If the latter is not available you can use n.d. or similar. The access date signals the date you accessed the information; it may or may not coincide with the publication date, and should be used independently of the latter whenever it is appropriate. You may also want to add a link note outside the citation such as [continuously updated source], especially if you have used a no-date date such as "n.d.". 50.75.226.250 (talk) 16:11, 30 April 2022 (UTC)
 * But what is the argument against since? For example, this page is used in 2022 Ukrainian refugee crisis for the number of Ukrainian refugees that reached France. The page was published on April 22, but the actual number of refugees gets updated every couple of days. Currently the following text is shown:
 * I am proposing a way to show this text instead,
 * in which “22 April 2022” becomes “since 22 April 2022”. --Grufo (talk) 19:19, 30 April 2022 (UTC)
 * Although editors have a choice of formats for the publication date, they have no choice for the date itself. You must use the publication date that appears in the source. "22 April" is one date (type:exact). "Since 22 April" is a different date (type:open range). If the source uses the exact date (as in this case) then you must use that date, as this is how the work will be best discovered. The access-date will indicate which version was actually accessed. If you want to "freeze" that version in the resulting citation, you should preemptively archive the page and add the archive information and relevant parameters to the citation, so that the archived version would be linked by default. An additional option is to present the source's fuzziness in wikitext, e.g. "according to frequently updated information, the number of heads on a pin was between zero and infinity on (citation-access-date)." 64.18.11.64 (talk) 22:57, 30 April 2022 (UTC)
 * The access-date parameter will indicate which version was accessed, but will not indicate that the page is constant update and different versions exist by design. Instead, since – i.e. “(since 22 April 2022)” – will indicate that the page was created exactly on 22 April with the explicit design of being constantly updated. --Grufo (talk) 14:33, 1 May 2022 (UTC)
 * Citations are not concerned with bibliographic information such as the design characteristics of the source, unless they directly impact discovery. In the example you provided, the source has an exact publication date. That is the date that the source will subsequently be classified with, and therefore the date by which it will most efficiently be found. There were several options given above to add the update information outside the citation, where it belongs: a link note and/or explanatory wikitext in the body or a footnote. These, combined with archive-url, archive-date, access-date and the implied default value of url-status will both give the update information and capture the version the article requires. 65.88.88.201 (talk) 15:51, 1 May 2022 (UTC)
 * What you propose as an alternative to since does not provide the reader with the information that the referenced page is in constant update. Newspapers often have “Live updates” pages, and these are usually treated differently (for example, the information that the page is in constant update normally accompanies the link to it). I understand that this is a new habit, born after internet was created, but this is precisely the time we live in. --Grufo (talk) 16:30, 1 May 2022 (UTC)
 * I think you may have misunderstood. The purpose of the citation is to provide the source of the material the editor used in wikitext, so that the reader can verify it. The fact that the source is dynamic has nothing to do with it. The editor is supposed to cite the version accessed. This is presently covered by the methods described above. 65.88.88.201 (talk) 16:43, 1 May 2022 (UTC)
 * The purpose of a reference is to provide a source, together with its context (is it a book? a newspaper? a conference? a constantly updated webpage?). What I do fail to understand is what kind of inconvenience the since parameter would create according to you and why you oppose it. --Grufo (talk) 04:16, 2 May 2022 (UTC)
 * The context is evident from the source one uses, and citations do not need to make it explicit. You can do this in accompanying text, if it is needed e.g. "according to CNN..." etc. Citations in Wikipedia exist to satisfy WP:V. The source information included must satisfy one requirement: it must help in some way to discover the source. Anything else is inconvenient, as clutter. Sources, especially continuing resources are often classified by date, and that is one way to be found. In my experience, I have not seen ever a date field anywhere that accepts "since" as part of the date. 50.75.226.250 (talk) 16:54, 2 May 2022 (UTC)
 * If I came here proposing this solution, maybe there are situations where the context is not evident at all, as shown also by . This page says that it was published on 22 April and that in France there are 51 375 refugees. The problem is that some days ago the same page published on 22 April said that the number of refugees was less than 50 000. The page does not say anywhere that it is constantly updated; I found out only by visiting it again after some days. Once again, you fail to explain why you oppose a since parameter, and limit yourself only to saying that you oppose it (“period”). P.S. “(since 22 April)” is a date, and does qualify as a means for classifying sources, if needed. --Grufo (talk) 20:54, 2 May 2022 (UTC)
 * I support this proposal. There are sources that I use that explicitly say to use "since DATE", when I am forced to use a manual citation in order to respect their statement. (E.g. the Angiosperm Phylogeny Website – see here.) Peter coxhead (talk) 14:42, 1 May 2022 (UTC)
 * Not a good example. There are several things to note here:
 * Note that the offered citation format is about the work (website), not a specific webpage (in-work location). Notice also that the update info is in [brackets], as an editor interpolation. The webpage itself (the in-work location) has an update date, just before the "Introductory" section
 * CS1/2 uses a different format, in which the version publication date (July 2017) should be used as the publication date of the website. The value originally published 2001 can be used, but it is not necessary. You may also add the info re:updating in that value. However, since the update date for the in-work location is given, and since presumably the wikitext depends on the updated information, then I would use the update date as publication date, and as you are citing a specific "edition" (version 14), I would probably use the version's date (July 2017) as the orig-date. Some of the acrobatics could be avoided by the use of pre-emptive archiving as discussed above.
 * The bottom line is, can the source be easily found with the way CS1/2 formats citations now? And what is the relative need and cost of adding yet another field, or of expanding the allowable date formats? 65.88.88.201 (talk) 16:47, 1 May 2022 (UTC)
 * Note that the offered citation format is about the work (website), not a specific webpage (in-work location). Notice also that the update info is in [brackets], as an editor interpolation. The webpage itself (the in-work location) has an update date, just before the "Introductory" section
 * CS1/2 uses a different format, in which the version publication date (July 2017) should be used as the publication date of the website. The value originally published 2001 can be used, but it is not necessary. You may also add the info re:updating in that value. However, since the update date for the in-work location is given, and since presumably the wikitext depends on the updated information, then I would use the update date as publication date, and as you are citing a specific "edition" (version 14), I would probably use the version's date (July 2017) as the orig-date. Some of the acrobatics could be avoided by the use of pre-emptive archiving as discussed above.
 * The bottom line is, can the source be easily found with the way CS1/2 formats citations now? And what is the relative need and cost of adding yet another field, or of expanding the allowable date formats? 65.88.88.201 (talk) 16:47, 1 May 2022 (UTC)
 * The bottom line is, can the source be easily found with the way CS1/2 formats citations now? And what is the relative need and cost of adding yet another field, or of expanding the allowable date formats? 65.88.88.201 (talk) 16:47, 1 May 2022 (UTC)


 * 'Since' just makes no sense. Since what? Since when? What since? What does it even mean to have say [Reference, Since 2009]? &#32; Headbomb {t · c · p · b} 08:56, 2 May 2022 (UTC)
 * The proposed parameter is not warranted because editors adding citations will usually not be able to determine when the website, or the portion of the website providing the information that supports the Wikipedia article claim, began operation. Even if it is possible, by the time the reader goes to the page to confirm the accuracy of the Wikipedia article or obtain more details the information about when website operation began may no longer be present or may be very difficult to find. The purpose of the citation is not to describe the source for the benefit of the reader who doesn't intend to read the source; it is for the benefit of the reader who does intend to read the source. Jc3s5h (talk) 14:50, 2 May 2022 (UTC)
 * Your argument is not very clear to me. Currently Wikipedia does not forbid to use “live” sources (like this one, for example). It seems you are suggesting that the since parameter will make it difficult to find an information due to it coming from a live source, and therefore the parameter should not be implemented. The problem is that not implementing the since parameter will not prevent editors from linking live sources – we already do that and the since parameter does not exist yet – but instead will only prevent editors from describing the source as “actively updated”. Both with and without a since parameter, the only way to make sure that an information remain easily findable is to use the archive-date parameter. --Grufo (talk) 15:54, 2 May 2022 (UTC)

Template:Cite AV media: more params to match Cite episode ?
For stuff that airs on TV that is not part of a episode of a show, it seems like there should be additional parameters for Cite AV, to allow specifying airdate and network. If one is referencing a TV special, a telefilm documentary, or such, it would seem that network and airdate would be useful to have around. -- 65.92.246.142 (talk) 02:06, 4 May 2022 (UTC)


 * We also need the parameter "via" added to this cite template., can something be done about this? Kailash29792 (talk)  06:04, 4 May 2022 (UTC)
 * Already supported:
 * —Trappist the monk (talk) 11:31, 4 May 2022 (UTC)
 * What of "airdate" and "network" ? -- 65.92.246.142 (talk) 22:44, 6 May 2022 (UTC)
 * —Trappist the monk (talk) 11:31, 4 May 2022 (UTC)
 * What of "airdate" and "network" ? -- 65.92.246.142 (talk) 22:44, 6 May 2022 (UTC)

standard templates question
Hello. I'm sure these things are set in stone since we are 20+ years in to this project, but I'm curious if there have ever been attempts to add a field for "Volume" to the standard book citation template, or a field for "Section" to the standard news template. A suggestion was made to use "Department" for "Section", although "Department", to me, has more to do with the organization of a newspaper as a business entity, and not as much to do with the organization of the physical newspaper itself. Thank you! Caro7200 (talk) 23:35, 4 May 2022 (UTC)
 * For attempts to add a field for "Volume" to the standard book citation template: Yes, and the attempt was successful:
 * The rendering has, of course, changed over time.
 * The department option came from me at a discussion at . Another option I mentioned was use to combine section and page enumerators in at.
 * I don't recall any specific discussion about making section be anything other than an alias of chapter.
 * —Trappist the monk (talk) 00:29, 5 May 2022 (UTC)
 * "Department" ... has more to do with the organization of a newspaper as a business entity, and not as much to do with the organization of the physical newspaper itself. This seems correct, but for citation purposes the terminology is interchangeable in news sources. However as a field name "section" may be ambiguous. Does it refer to paginated sections, e.g. "Sports" section pp. x–y? A section in a page, e.g. "Weather" p. x col. y? A titled section in an article e.g. "X did Y", § "Reaction from Z"? A bunch of paragraphs in an article, e.g. "A bashes B" ¶¶ 2–3? 71.247.146.98 (talk) 12:17, 5 May 2022 (UTC)
 * I don't recall any specific discussion about making section be anything other than an alias of chapter.
 * —Trappist the monk (talk) 00:29, 5 May 2022 (UTC)
 * "Department" ... has more to do with the organization of a newspaper as a business entity, and not as much to do with the organization of the physical newspaper itself. This seems correct, but for citation purposes the terminology is interchangeable in news sources. However as a field name "section" may be ambiguous. Does it refer to paginated sections, e.g. "Sports" section pp. x–y? A section in a page, e.g. "Weather" p. x col. y? A titled section in an article e.g. "X did Y", § "Reaction from Z"? A bunch of paragraphs in an article, e.g. "A bashes B" ¶¶ 2–3? 71.247.146.98 (talk) 12:17, 5 May 2022 (UTC)
 * "Department" ... has more to do with the organization of a newspaper as a business entity, and not as much to do with the organization of the physical newspaper itself. This seems correct, but for citation purposes the terminology is interchangeable in news sources. However as a field name "section" may be ambiguous. Does it refer to paginated sections, e.g. "Sports" section pp. x–y? A section in a page, e.g. "Weather" p. x col. y? A titled section in an article e.g. "X did Y", § "Reaction from Z"? A bunch of paragraphs in an article, e.g. "A bashes B" ¶¶ 2–3? 71.247.146.98 (talk) 12:17, 5 May 2022 (UTC)

Discussion regarding Citation bot's formatting of various citation templates
There is currently a discussion at the Citation bot's talk page in regards to how the bot/tool formats certain websites/content publishers within specific citation templates. The discussion can be found here: User talk:Citation bot. - Favre1fan93 (talk) 21:55, 7 May 2022 (UTC)

How do I use a "|" within a cite web template
I am trying to use a "|" character in a cite web template. It is causing problems I have tried to wrap it in "" – " which undoes the break but does not present cleanly. See Duncan Robinson (basketball).-TonyTheTiger (T / C / WP:FOUR / WP:CHICAGO / WP:WAWARD) 16:33, 8 May 2022 (UTC)


 * Izno (talk) 16:47, 8 May 2022 (UTC)
 * Locate what it is that you want to cite and then use page. The writers of "Michigan Basketball History & Records" couldn't be bothered to use the 'Section &lt;nn> | &lt;title>' headings anywhere but on the first page so those 'headings' are useless as search terms.  Fortunately the document is paginated so use page or pages and ignore the headings.
 * is not a template but is.
 * —Trappist the monk (talk) 16:54, 8 May 2022 (UTC)
 * I think you are mistaken re "couldn't be bothered": the link in this citation goes only to a single section of a larger document. So the reason the other section headings do not appear is that the other sections themselves do not appear. A vertical bar within a title is usually a mistake caused by software that automates the conversion of urls to citations. Splitting it out into separate pieces of metadata in separate parameters and standardizing the punctuation is usually a better choice. Because this is a part of a larger work, it is reasonable in this case to use title for the part and work for the larger work:
 * —David Eppstein (talk) 19:31, 8 May 2022 (UTC)
 * Thx-TonyTheTiger (T / C / WP:FOUR / WP:CHICAGO / WP:WAWARD) 21:12, 8 May 2022 (UTC)
 * —David Eppstein (talk) 19:31, 8 May 2022 (UTC)
 * Thx-TonyTheTiger (T / C / WP:FOUR / WP:CHICAGO / WP:WAWARD) 21:12, 8 May 2022 (UTC)
 * —David Eppstein (talk) 19:31, 8 May 2022 (UTC)
 * Thx-TonyTheTiger (T / C / WP:FOUR / WP:CHICAGO / WP:WAWARD) 21:12, 8 May 2022 (UTC)

-TonyTheTiger (T / C / WP:FOUR / WP:CHICAGO / WP:WAWARD) 21:12, 8 May 2022 (UTC)

Number and issue
Is there a way to display both the issue in volume and the overall number, of a publication? I'm getting a CS1 error by specifying this. -- 65.92.246.142 (talk) 22:48, 6 May 2022 (UTC)
 * Please explain what you are trying to cite. 63.117.211.42 (talk) 01:59, 7 May 2022 (UTC)
 * Whoops, fixed the missing journal line -- 65.92.246.142 (talk) 02:50, 7 May 2022 (UTC)
 * This is the 212th publication of this journal, which occurs as volume 9 number 4, published on 1815-09-23. I would think there should be a way to indicate the overall number (no. 212) as well as the number in volume (no.4 of vol.9) -- 65.92.246.142 (talk) 03:05, 7 May 2022 (UTC)
 * I think in library science terminology the 212 is called a "whole number". More about whole numbering. According to that source, it's a good idea to provide both if available to allow matching more easily. If we did support whole numbering it would need to be indicated as being a whole number. It is implied if no volume. In this case there is both volume+number and a whole number. --  Green  C  03:37, 7 May 2022 (UTC)
 * Let's keep in mind that citations in Wikipedia are not bibliographic records, and not all bibliographic systems may use the whole number. Library systems may also use their own classifications, and not all of these may use a whole number. The issue in question can be discovered easily with the available CS1\2 facility, the combination date+volume/issue. There are several OCLC records for this source, including, which provides links to the magazine's archive at IA, uploaded by the Boston Public Library. The specific in-source location is here: in-source location. Use cite magazine, and do not use the "whole number", which is mostly for the use of librarians. 50.75.226.250 (talk) 15:06, 7 May 2022 (UTC)
 * You could hack it a bit if you feel that both items are useful:
 * Or something. – Jonesey95 (talk) 14:11, 8 May 2022 (UTC)
 * That is incorrect use. The "volume" parameter should only have volume information, not whatever the editor thinks it should have. Secondly, and more important, it is confusing to readers. What does it even mean? Is this the 4th issue of the 9th volume? If the issue number refers to a magazine issue number, what does the other number stand for? Is the issue actually 4 or is it 212? Why is it there in the first place, and should I use it to discover the source? Or use both? Will it help? Confusion. Citations are there for readers, not editors, and readers are not supposed to agonize over them. It is interesting that this simple fact has to be pointed out again and again and again. 65.88.88.201 (talk) 16:16, 8 May 2022 (UTC)
 * Showing both would help in finding the resource, should one or the other be used but not the other -- 65.92.246.142 (talk) 22:14, 8 May 2022 (UTC)
 * Re "What does it even mean?", you would have to ask the magazine's publisher. If the issue states that it is volume 9, issue 4, number 212, then that information may be needed to locate the issue on a library shelf or database, or to verify by looking at the magazine cover that you have the issue you seek. Citations exist to help readers verify information by finding the cited sources. – Jonesey95 (talk) 03:42, 9 May 2022 (UTC)
 * The particular source is classified by volume and issue number . The main bibliographic provider for this publication seems to be the US Library of Congress, their catalog entry is here. Notice that the secondary numbering ("Whole number" and alternate volume/issue numbers) are in a "Description" field. This format is used by several other downstream metadata providers supplied by the LoC. The New York Public Library, another primary biblio provider, has the additional numbering in a "Numbering notes" field in its Research catalog (here, scroll to "Details"). The Boston Public Library, which uploaded the publication to Internet Archive as authoritative scans, adds that info in a "Notes" field (here, click on "Full details"). Most downstream metadata providers whose product is freely accessible by Wikipedia readers (via Amazon searches, Google searches etc) do not index these fields for quick retrieval. If they do, they are secondary indices whose search results may appear in the first results page or the 34th. The descriptions and notes themselves are terse and likely confusing to the average reader. They are there for people whose job is to retrieve this information for lay readers, when the lay person may know any of these numbers without knowing "better" information that is always indexed: the date, article title, author, or issue number+volume. So the additional numbering info is of secondary value in discovering the source. However, all of the above are moot, because the source exists online on the web, at a stable repository, uploaded by a reliable provider as a free resource. Which makes this to be included in any citation of the source. That is how the reader will find it fast to verify the article wikitext.
 * Including the secondary information in a citation without further explanation is confusing: the proposed "volume 9, issue 4, no. 212" could mean anything. Perhaps volume 9 has been issued 4 times and includes 212 (or more) issues? Does the "no." refer to the 212th part of issue 4 instead? To disentagle all that an outside note would be required. But the editor should fill in the volume and issue fields as they are most likely to be found: 9 4. 50.75.226.250 (talk) 16:33, 9 May 2022 (UTC)
 * If you examine the actual cover page, it includes all 3 numbers, the volume, the issue, the whole number. It is not an exercise in mathematics figuring out what the whole number is. It is published right on the printed page. So, I would expect that some people might just classify it that way -- 65.92.246.142 (talk) 00:06, 11 May 2022 (UTC)
 * If you examine the actual cover page, it includes all 3 numbers, the volume, the issue, the whole number. It is not an exercise in mathematics figuring out what the whole number is. It is published right on the printed page. So, I would expect that some people might just classify it that way -- 65.92.246.142 (talk) 00:06, 11 May 2022 (UTC)

Adding "via" to Template:Cite AV media
How do we do it to the cite template's TemplateData? Kailash29792 (talk)  09:40, 3 May 2022 (UTC)


 * @Kailash29792 ✅ by in . GoingBatty (talk) 03:14, 12 May 2022 (UTC)

chapter in cite conference
The existence of chapter in cite conference should be an error that CS1/2 tracks. AManWithNoPlan (talk) 20:47, 11 May 2022 (UTC)
 * That parameter is used in roughly 500 pages. Are they all errors? It appears that chapter and title are interpreted as title and book-title, which is a clever bit of processing and makes sense, given that conference proceedings are essentially a book. Here's one that appears to be working fine:
 * That is from Abstract data type. – Jonesey95 (talk) 02:07, 12 May 2022 (UTC)
 * Not all, but some. When chapter title and book-title are all set, then error. AManWithNoPlan (talk) 20:26, 13 May 2022 (UTC)

Custom note
How to add custom note if user has to click something to access content? Should I use "format" parameter? Eurohunter (talk) 16:28, 13 May 2022 (UTC)
 * Don't misuse parameters. format is to be used to identify the electronic file format of the associated url; nothing else.  To add a note:
 * —Trappist the monk (talk) 16:40, 13 May 2022 (UTC)
 * I think there should be separate parameter for custom note. Eurohunter (talk) 15:46, 14 May 2022 (UTC)
 * This is not a new idea but it has not gotten sufficient support to be implemented. Here are two previous conversations:
 * There may be other discussions in the archives that I didn't find.
 * —Trappist the monk (talk) 16:36, 14 May 2022 (UTC)
 * There may be other discussions in the archives that I didn't find.
 * —Trappist the monk (talk) 16:36, 14 May 2022 (UTC)
 * —Trappist the monk (talk) 16:36, 14 May 2022 (UTC)

Author-link
Why can't author names which are presented in normal order, same as our article titles, not be wikilinked? The instruction says:

I always assumed that author-link was used when the template was being used as a alphabetical source link so the author names were presented surname first, and so couldn't be linked. But that it was OK to wikilink when the template is being used for numbered citations and so the author parameter is used to present the name in normal order, same as our article titles. The result appears to be the same - a wikilink takes the reader to the article, and author-link takes the reader to the article. So, does author-link do something extra that I'm not aware of? SilkTork (talk) 17:02, 14 May 2022 (UTC)
 * It is permissible to wikilink author&lt;n:
 * The "Do not wikilink..." text was added to the documentation at this edit for a confused editor. Perhaps Editor Jonesey95 can clarify...
 * —Trappist the monk (talk) 17:16, 14 May 2022 (UTC)
 * Curiously, Trappist the monk, it was one of your edits that prompted me to look into the author-link situation - you delinked a wikilink in order to add an author-link: . I wondered why you did that, but rather than troubling you, came to the template documentation to see what the rule was. According to the rule, what you did was correct. I assume the edit was semi-automatic. SilkTork (talk) 22:02, 14 May 2022 (UTC)
 * I looked through the archives of this talk page and my own talk page in the date range surrounding that edit (Feb 2020), and I did not find a relevant discussion. I do not know why I added that text. I see that many parallel parameters have the same admonition. – Jonesey95 (talk) 22:22, 14 May 2022 (UTC)
 * For last (assuming that the value assigned to last is a family name) and first, the 'do-not-wikilink' wording is appropriate. For author? Not so much, I think.  If the source author is a corporate name or mononym, wikilinking that name is acceptable.  Because we have historically equated last with author as full and complete aliases, those parameters get identical handling in Module:Citation/CS1.  Over time, we have become more strict on wikilinking first so we might consider enforcing the do-not-wikilink for last and surname and similarly for contributor-last, editor-last, interviewer-last, translator-last...
 * Or somehow change the documentation so that it better describes the intent of author v. last.
 * —Trappist the monk (talk) 23:03, 14 May 2022 (UTC)
 * Well, yes and no. You changed this:
 * to this:
 * which, as you can see, causes Module:Citation/CS1 to emit an error message. It was that error message that attracted my attention so I changed your edit by removing the wikilink in Geoffrey Chaucer and by moving the existing Geoffrey Chaucer so that is followed author, changed publisher to website, added archive-url and archive-date, and fixed title.
 * The edit was a manual edit augmented by a cleanup script that converted curley quotes to straight quotes.
 * —Trappist the monk (talk) 22:42, 14 May 2022 (UTC)
 * Cool, that explains it. My own edit was prompted by the previous edit which had replaced the natural name author parameter with a reversed author name plus adding the author-link parameter. I hadn't tidied up properly by removing the author-link parameter, so thanks for doing that.
 * "Or somehow change the documentation so that it better describes the intent of author v. last" - that makes sense. last is applicable only when the template is being used to create an alphabetical list of authors used as sources, such lists being placed after the references section; while author is applicable when the template is being used to create a numbered list of sources for the references section. I don't think this usage is made clear, and quite commonly  last is used for all citations, so authors' names are frequently reversed in the numbered references section as well as the alphabetical list, so the author names in the numbered list read awkwardly; and it doesn't make sense because the numbered list is not alphabetical, so there is no point in reversing people's names. As an existing error it is somewhat tolerable, but when editors deliberately change author to last it becomes irksome. SilkTork (talk) 10:42, 15 May 2022 (UTC)
 * That is not a correct understanding. We prefer but do not require the use of the last/first parameters for multiple reasons, wherever they appear. The author parameter should be used almost exclusively when last would be an inappropriate description, such as for organizational names. Izno (talk) 17:36, 15 May 2022 (UTC)
 * Agree with Izno's reply. Two additional points: short references use the last name as common practice, when there is one (there are uncommon exceptions). The related long citation list is very often arranged by alpha. Secondly, most listings of citable material, whether these are marketing lists, repository lists, library lists etc. mainly arrange/index their lists by author's last name if there is one. By offering that information first, a Wikipedia citation signals the reader the easiest/fastest way to find the source. 68.173.78.83 (talk) 18:12, 15 May 2022 (UTC)
 * which, as you can see, causes Module:Citation/CS1 to emit an error message. It was that error message that attracted my attention so I changed your edit by removing the wikilink in Geoffrey Chaucer and by moving the existing Geoffrey Chaucer so that is followed author, changed publisher to website, added archive-url and archive-date, and fixed title.
 * The edit was a manual edit augmented by a cleanup script that converted curley quotes to straight quotes.
 * —Trappist the monk (talk) 22:42, 14 May 2022 (UTC)
 * Cool, that explains it. My own edit was prompted by the previous edit which had replaced the natural name author parameter with a reversed author name plus adding the author-link parameter. I hadn't tidied up properly by removing the author-link parameter, so thanks for doing that.
 * "Or somehow change the documentation so that it better describes the intent of author v. last" - that makes sense. last is applicable only when the template is being used to create an alphabetical list of authors used as sources, such lists being placed after the references section; while author is applicable when the template is being used to create a numbered list of sources for the references section. I don't think this usage is made clear, and quite commonly  last is used for all citations, so authors' names are frequently reversed in the numbered references section as well as the alphabetical list, so the author names in the numbered list read awkwardly; and it doesn't make sense because the numbered list is not alphabetical, so there is no point in reversing people's names. As an existing error it is somewhat tolerable, but when editors deliberately change author to last it becomes irksome. SilkTork (talk) 10:42, 15 May 2022 (UTC)
 * That is not a correct understanding. We prefer but do not require the use of the last/first parameters for multiple reasons, wherever they appear. The author parameter should be used almost exclusively when last would be an inappropriate description, such as for organizational names. Izno (talk) 17:36, 15 May 2022 (UTC)
 * Agree with Izno's reply. Two additional points: short references use the last name as common practice, when there is one (there are uncommon exceptions). The related long citation list is very often arranged by alpha. Secondly, most listings of citable material, whether these are marketing lists, repository lists, library lists etc. mainly arrange/index their lists by author's last name if there is one. By offering that information first, a Wikipedia citation signals the reader the easiest/fastest way to find the source. 68.173.78.83 (talk) 18:12, 15 May 2022 (UTC)
 * Agree with Izno's reply. Two additional points: short references use the last name as common practice, when there is one (there are uncommon exceptions). The related long citation list is very often arranged by alpha. Secondly, most listings of citable material, whether these are marketing lists, repository lists, library lists etc. mainly arrange/index their lists by author's last name if there is one. By offering that information first, a Wikipedia citation signals the reader the easiest/fastest way to find the source. 68.173.78.83 (talk) 18:12, 15 May 2022 (UTC)

Make "original" unclickable for archived dead links?
I'm frequently confused when I quickly want to access an archived link. If the link is (likely) dead, I can click the second link in the citation ("Archived"), and if the link is live, I want to click the first link (not "original"). Example live link with archive: Example dead link with archive

I often find myself clicking on the 'original', which leads to the dead link. It's very rarely a link anybody would want to click, so why give them the option?

The Italian Wikipedia has a better display here in my opinion, with something like hover title, which prevents you from clicking the url but still displays it: (or see an example on itwiki)

Do other people have the same issue? And would this be an improvement here too?

Femke (talk) 19:24, 14 May 2022 (UTC)

When hovering over a link, it shows up in the lower bar of the browser. I find it clear and consistent. The Italian way is interesting, but changes the location of where to look during hover which adds a new level of complexity. It also makes it hard to copy-paste the URL with right-click-copy-link which is useful at times. I think the main reason for the Italian way is to prevent users from clicking on the presumed dead link. But it's not always accurate the primary link is dead - sometimes it is incorrectly marked dead, sometimes they come back alive - by hiding behind a hover it makes it less likely this problem will be detected and fixed. -- Green  C  21:43, 14 May 2022 (UTC)

In cases of pre-emptive archiving, both links are live, and the default link is the archive link. This way, when/if the original link becomes unavailable there will be no effect on verifiability. In the meantime, it is not a good idea, for both semantic and presentation reasons, to replace the still-live original with a non-linking URL. 68.173.78.83 (talk) 00:29, 15 May 2022 (UTC)

Cite letter with no author
Cite letter is based on Cite press release, anyway the absence of an author (optional according to template documentation) returns a maintenance error, even if the latter template suggests |author=. Is there a workaround? Thanks in advance.--Carnby (talk) 09:41, 15 May 2022 (UTC)

The documentation is confusing, as it shows "first" and "last" among the required patameters, but as optional in the TemplateData rendition. I would use author (or Anonymous), if it works. In general, letters need attention. They should be treated as primary sources with the corresponding reliability & notability issues. Their publisher is very important, for the same reasons. 2603:7000:2B42:BB00:ED36:A005:6AE4:C278 (talk) 13:47, 15 May 2022 (UTC)


 * Your suggestion worked, thank you.--Carnby (talk) 16:45, 15 May 2022 (UTC)