Help talk:Citation Style 1/Archive 58

subscription=yes
The subscription=yes was deprecated this year. Maybe it's only me, but I don't get how the five (?) new &hellip;-access=&hellip; parameters are supposed to work. The general situation is a site using JavaScript to detect an AdBlock and not working at all, if JavaScript is disabled. What is this, &hellip;-access=limited or &hellip;-access=subscription ? Sometimes I'm too lazy to disallow JS, and get either the complete page with some anti-AdBlock-Ad, a part of the page, or nothing. If I only wanted to verify a reference a part can be already good enough, but maybe a critical BLP detail is not covered in the visible part. In that case I added subscription=yes and ignored the issue (if possible per BLP policy, no wild and wonderful statements.) Today I tested article-url-access=limited, epic fail, I got confusing error messages about a chapter-url-access=&hellip; , and I have no clue what this is: It is certainly not mentioned on Help:Citation Style 1, it is not explained on Template:Cite web/doc, and the cross-namespace redirection of Template talk:Cite web to Help talk:Citation Style 1 used to be a speedy deletion reason about 12 years ago. please help. –84.46.53.102 (talk) 03:42, 23 June 2019 (UTC)
 * There's six of them, but it's unclear what they are for. The docs say If the restriction applies to an identifier, these parameters should be omitted.Since the docs say citations within a given article should follow a consistent style, it looks like the access icons need to be suppressed throughout the article.   Hawkeye7   (discuss)  05:41, 23 June 2019 (UTC)
 * Neither of your two diffs show any epic fail. I do see that  caused an error message.  article-url-access is an alias of chapter-url-access in the same way that article-url is an alias of chapter-url.  That error message occurs because there is no article-url in the citation template, which, in any case, is not supported by.
 * I get that the 'chapter' error message is a bit confusing. That issue has been fixed in Module:Citation/CS1/sandbox:
 * —Trappist the monk (talk) 10:39, 23 June 2019 (UTC)
 * Thanks, I'll test article-url-access=&hellip; again when I stumble over a reference, where nothing is wrong with the URL, but major parts of the article are blurred/hidden, unless "whatever" (registration, subscription, free trial countdown exhausted, no AdBlock, 3rd party cookies required, &hellip;) –84.46.53.196 (talk) 21:02, 24 June 2019 (UTC)
 * —Trappist the monk (talk) 10:39, 23 June 2019 (UTC)
 * Thanks, I'll test article-url-access=&hellip; again when I stumble over a reference, where nothing is wrong with the URL, but major parts of the article are blurred/hidden, unless "whatever" (registration, subscription, free trial countdown exhausted, no AdBlock, 3rd party cookies required, &hellip;) –84.46.53.196 (talk) 21:02, 24 June 2019 (UTC)
 * —Trappist the monk (talk) 10:39, 23 June 2019 (UTC)
 * Thanks, I'll test article-url-access=&hellip; again when I stumble over a reference, where nothing is wrong with the URL, but major parts of the article are blurred/hidden, unless "whatever" (registration, subscription, free trial countdown exhausted, no AdBlock, 3rd party cookies required, &hellip;) –84.46.53.196 (talk) 21:02, 24 June 2019 (UTC)

Here's my problem:
 * How do you add the "subscription required"?  Hawkeye7   (discuss)  23:51, 23 June 2019 (UTC)
 * Like this:
 * in its present form doesn't bring anything to the table. It might if it were rewritten to take advantage of features available in  – notably url-accessn.  But, with only one identifier, not needed here.
 * —Trappist the monk (talk) 00:35, 24 June 2019 (UTC)
 * in its present form doesn't bring anything to the table. It might if it were rewritten to take advantage of features available in  – notably url-accessn.  But, with only one identifier, not needed here.
 * —Trappist the monk (talk) 00:35, 24 June 2019 (UTC)
 * —Trappist the monk (talk) 00:35, 24 June 2019 (UTC)

Zbl check
As a follow up on Help talk:Citation Style 1/Archive 36 and Help talk:Citation Style 1/Archive 39, the Zbl error checking should allow for all numeric (8 digit specifically) possibilities, as Zbl can have temporary assignments, such as and  found in Vladimir Mazya, or  found in Lou van den Dries.

Those could be put in a Category:CS1 maintenance: Temporary Zbl or similar. &#32; Headbomb {t · c · p · b} 17:51, 24 June 2019 (UTC)
 * I have discovered that part of the code already has a maint cat that has fortunately never been used. The code looks for a 'zbl' prefix in the identifier: Zbl 1260.11001.  When found, the prefix is stripped and the article added to an undefined category.  Because undefined, if ever a zbl had the prefix we would have gotten a glaring red script error.  That we haven't (or that no one has complained), perhaps this check is unnecessary.  For the moment the check remains in place:
 * and a sanity check using the zbl without prefix:
 * Assuming that temporary assignments are always eight digits:
 * but seven digits:
 * —Trappist the monk (talk) 12:34, 25 June 2019 (UTC)
 * ... I don't like that. The error should be pointed out. This isn't like in the case of the PMCID where the 'pmc' is actually part of the identifier. &#32; Headbomb {t · c · p · b} 15:46, 25 June 2019 (UTC)
 * Tweaked.
 * —Trappist the monk (talk) 16:40, 25 June 2019 (UTC)
 * but seven digits:
 * —Trappist the monk (talk) 12:34, 25 June 2019 (UTC)
 * ... I don't like that. The error should be pointed out. This isn't like in the case of the PMCID where the 'pmc' is actually part of the identifier. &#32; Headbomb {t · c · p · b} 15:46, 25 June 2019 (UTC)
 * Tweaked.
 * —Trappist the monk (talk) 16:40, 25 June 2019 (UTC)
 * ... I don't like that. The error should be pointed out. This isn't like in the case of the PMCID where the 'pmc' is actually part of the identifier. &#32; Headbomb {t · c · p · b} 15:46, 25 June 2019 (UTC)
 * Tweaked.
 * —Trappist the monk (talk) 16:40, 25 June 2019 (UTC)

Nomination for deletion of Template:Cs1 function
Template:Cs1 function has been nominated for deletion. You are invited to comment on the discussion at the template's entry on the Templates for discussion page. – Finnusertop (talk ⋅ contribs) 19:08, 25 June 2019 (UTC)

url + title=none
For stylistic reasons, the title in needs suppressing. However, when you set none, you get

I believe in this case, we should have something like or even something like or instead. &#32; Headbomb {t · c · p · b} 22:52, 6 June 2019 (UTC)
 * Bach, Eric (2001), Discrete Dynamics in Nature and Society, 6 (2): 145–146,, available here.
 * Bach, Eric (2001), Discrete Dynamics in Nature and Society, 6 (2): 145–146,, available at https://downloads.hindawi.com/journals/ddns/2001/120718.pdf
 * Bach, Eric (2001), Discrete Dynamics in Nature and Society, 6 (2): 145–146,
 * Bach, Eric (2001), Discrete Dynamics in Nature and Society, 6 (2): 145–146,
 * Lenore Blum? That source is open access, there is a link to the pdf from the doi-linked publisher page so:
 * This form is similar to the Meer, Vavasis, and McNicholl citations in that  tag.
 * The  links above cause my current-version Chrome browser complain about security certificate mismatch.  The link at the doi-linked publisher's page is an   url.
 * —Trappist the monk (talk) 23:56, 6 June 2019 (UTC)
 * The  links above cause my current-version Chrome browser complain about security certificate mismatch.  The link at the doi-linked publisher's page is an   url.
 * —Trappist the monk (talk) 23:56, 6 June 2019 (UTC)
 * —Trappist the monk (talk) 23:56, 6 June 2019 (UTC)


 * Maybe so, but that doesn't detract from the general point that this is a feature that should be supported, e.g.
 * &#32; Headbomb {t · c · p · b} 21:58, 25 June 2019 (UTC)
 * &#32; Headbomb {t · c · p · b} 21:58, 25 June 2019 (UTC)

Monkbot
Now that a bot is changing field names, a concerning issue arises if "website=" can take no wiki-markup. "Website=" automatically italicizes anything in the field. Yet many things, like the names of TV networks (ABC, CNN) and non-periodical sites like Rotten Tomatoes and AllMusic, are not italicized, and having them appear non-italicized in text and italicized in References is inconsistent and contrary to most standard footnoting style. I would note the MOS indicates that non-periodical websites are not italicized and that only this template forces that. What can we do to address this? --Tenebrae (talk) 21:00, 27 June 2019 (UTC)
 * No. But see . --Izno (talk) 21:05, 27 June 2019 (UTC)
 * This conversation apparently derives from one begun (and now closed because this one exists) on my talk page:.
 * —Trappist the monk (talk) 22:16, 27 June 2019 (UTC)
 * —Trappist the monk (talk) 22:16, 27 June 2019 (UTC)

The bot is doing the right thing. When I cite something I read on https://www.nbc.com, I am not citing the network because the network is on television. I am citing the website which is a periodical and just so happens to share a name with the network. The publisher is "NBCUniversal". The "website=" is the proper parameter to use in this example. If you believe "website=" should not be italicized, let's have that discussion. Or advocate for another parameter for non-periodical websites (but I don't see why they should be treated differently. They are a body of work and should be italicized). Misusing "publisher=" is not a solution no matter how long that has been the status quo. Rotten Tomatoes is published by Fandango. AllMusic is published by RhythmOne. --- Coffee  and crumbs  04:24, 28 June 2019 (UTC)


 * Neither AllMusic nor Rotten Tomatoes is italicized, per WP:COMMONNAME and standard usage. And NBCUniversal is not a publisher but an owner. I don't believe Wiikipedia's credibility is helped by redefining words to try to make them fit an arbitrary footnote style. --Tenebrae (talk) 17:31, 28 June 2019 (UTC)
 * This nothing to do with COMMONNAME which is an article title policy. This is about references. Check MLA, APA, or any citation style. Websites are italicized. Please continue this discussion in the above RfC. --- Coffee  and crumbs  17:45, 28 June 2019 (UTC)

Add Introduction, Prologue, Foreword, Epilogue, Afterword?
I don't know how to do this myself, but would Wikipedia and/or fellow Wikipedians please consider how to add both "Introduction" and "Introduction-first"/"Introduction-last" to the book citation template? Ditto for Prologue, Foreword, Epilogue, Afterword? I ask because often one or more experts will kindly not edit a famous book but will add expert commentary before and/or after a book's text. Thank you Aboudaqn (talk) 18:28, 29 June 2019 (UTC)
 * What part of the book are you citing as a reference, and why does it not work for you to cite that specific part as a contribution within a book rather than requiring separate parameters? —David Eppstein (talk) 18:37, 29 June 2019 (UTC)
 * supports contribution and contributor (and firt/last variants):
 * This feature is only for (or book cites using ).
 * When the book's author or editor also wrote the preface, introduction, ... use chapter or contribution.
 * —Trappist the monk (talk) 19:30, 29 June 2019 (UTC)
 * This feature is only for (or book cites using ).
 * When the book's author or editor also wrote the preface, introduction, ... use chapter or contribution.
 * —Trappist the monk (talk) 19:30, 29 June 2019 (UTC)
 * —Trappist the monk (talk) 19:30, 29 June 2019 (UTC)

cite encyclopedia without |title=
Way back at this discussion → Help talk:Citation Style 1/Archive 4, I wrote code in Module:Citation/CS1 that remaps the various title, article (or alias), and encyclopedia parameters of and  templates according to whichever combination of parameters is present. One of those remappings is encyclopedia to title when title and article are both missing / empty. What that does is this:

I don't know what I was thinking. cs1|2 citations require titles.

Internally, Module:Citation/CS1 copies the value assigned to encyclopedia into a meta-parameter. When there is no title and no article, then the module copies the content of  to the title meta-parameter   and resets   to an empty string. This satisfies the requirement for a cs1|2 title. But should it?

I discovered this because I've been working on cleaning up various cs1 wrapper templates that link to wikisource (,, etc). Those templates force a title when wstitle and title are missing or empty:

Beyond the fact that there is no title, the problem with this way of dealing with the missing title is that the label portion of that wikilink, the part that goes into the citation's metadata, has html markup which it should not. So, I have tweaked the module to be more like other templates where title is missing or empty:

—Trappist the monk (talk) 15:41, 28 June 2019 (UTC)
 * Maybe you want to reconsider. Encyclopedias are composed of articles, entries, or contributions, and are indexed accordingly. There are no titled works present, apart from the encyclopedia title itself and perhaps, special chapter-titles. Just like uses title as an alias of the "work" (the book). It would make more sense to introduce (in both  and ) a parameter book and to alias it with title rather than increasing the confusion. Edited collections are also books, in print or otherwise. 98.0.246.242 (talk) 01:55, 29 June 2019 (UTC)
 * I don't think so. This change is not made to do anything more than alert editors that a  citation template is missing an article title whether that be title or any of the article aliases.  Restructuring the existing template suite to be more semantically correct may be something that we want to do but that is a topic for another time and place.
 * —Trappist the monk (talk) 09:15, 29 June 2019 (UTC)
 * And how do you cite the encyclopedia then? Is title=none an option? 98.0.145.210 (talk) 12:17, 29 June 2019 (UTC)
 * Why would you want to cite an entire encyclopedia instead of a relevant article in an encyclopedia?
 * —Trappist the monk (talk) 12:37, 29 June 2019 (UTC)
 * If you are including the encyclopedia in a bibliography, or are citing several articles from the same encyclopedia:Nigel Ish (talk) 12:53, 29 June 2019 (UTC)
 * — Preceding unsigned comment added by Nigel Ish (talk • contribs) 12:53, 29 June 2019 (UTC)
 * All of that repeated bibliographic detail ... (use )
 * You can do this:
 * The template is happy because it has a title; the metadata are happy because there is a title. In this form,  is mostly the same as  in both its rendering and its metadata.
 * —Trappist the monk (talk) 13:13, 29 June 2019 (UTC)
 * So, is title an alias for "work" (encyclopedia) in now? 65.88.88.75 (talk) 15:22, 29 June 2019 (UTC)
 * This is, it is different. In the discussion I mentioned, I discuss how Module:Citation/CS1 remaps the various parameters.  When title is empty or omitted, encyclopedia is promoted to title so that the citation's metadata will have that meaningful information .  Given title alone, nothing promotes and we have good metadata.  encyclopedia is  an alias of title nor is title an alias of article though in both cases the former can be promoted to the latter under appropriate conditions.
 * —Trappist the monk (talk) 16:20, 29 June 2019 (UTC)
 * This maybe the wrong forum. Questions involving basic human understanding of how citations are to be applied in plain everyday English (hopefully), which may be the common lot of the average Wikipedia editor, are answered with technicalities about keeping middleware happy. Why is that? Not asking sarcastically or with any intent to insult or belittle. 24.105.132.254 (talk) 18:51, 29 June 2019 (UTC)
 * Because I'm a technical sort of person, I write from a technical point of view. This is why I do not write the documentation for these templates.
 * —Trappist the monk (talk) 19:38, 29 June 2019 (UTC)
 * So, is title an alias for "work" (encyclopedia) in now? 65.88.88.75 (talk) 15:22, 29 June 2019 (UTC)
 * This is, it is different. In the discussion I mentioned, I discuss how Module:Citation/CS1 remaps the various parameters.  When title is empty or omitted, encyclopedia is promoted to title so that the citation's metadata will have that meaningful information .  Given title alone, nothing promotes and we have good metadata.  encyclopedia is  an alias of title nor is title an alias of article though in both cases the former can be promoted to the latter under appropriate conditions.
 * —Trappist the monk (talk) 16:20, 29 June 2019 (UTC)
 * This maybe the wrong forum. Questions involving basic human understanding of how citations are to be applied in plain everyday English (hopefully), which may be the common lot of the average Wikipedia editor, are answered with technicalities about keeping middleware happy. Why is that? Not asking sarcastically or with any intent to insult or belittle. 24.105.132.254 (talk) 18:51, 29 June 2019 (UTC)
 * Because I'm a technical sort of person, I write from a technical point of view. This is why I do not write the documentation for these templates.
 * —Trappist the monk (talk) 19:38, 29 June 2019 (UTC)

Dates in author field
Hello, I have located around 200 author fields that contain a date of some form prefixed by "Published". Would be good if someone could move the date detail to the date field, assuming the field is blank, formatting it appropriately. If the date matches the date in the author field then ignore it. In either case just blank the author field. Hopefully that should just leave a few cases to handle manually where the date from the author field does not match the date field. Keith D (talk) 19:59, 29 June 2019 (UTC)

bibcode/ADS - URL base is deprecated
The cite journal template needs to be updated. I couldn't find which particular page codes for the bibcode= parameter in cite journal, so sorry if this is the wrong talk page. See for what needs doing. For example,
 * the bibcode parameter in Template:cite journal gives http://adsabs.harvard.edu/abs/1917SPAW.......142E, which is deprecated (the URL, not the article; Lambda is part of today's standard cosmological model, even though the article only officially has 10 citations...);
 * Template:bibcode now (01:55, 30 June 2019 (UTC)) gives https://ui.adsabs.harvard.edu/abs/1917SPAW.......142E, which is recommended (unfortunately it's extremely heavy in terms of third-party scripts, including GAFAM stuff like Google Analytics; uMatrix provides an efficient way of separating necessary third-party scripts from privacy-violating ones).

The fix will affect a huge amount of pages. Double-check and triple-check before hitting "Publish changes"...! Boud (talk) 01:55, 30 June 2019 (UTC)
 * We have already discussed this here and here. Please search before making a new thread in the future. --Izno (talk) 13:27, 30 June 2019 (UTC)
 * I reverted the change at Template:Bibcode, the javascript version load times for abstract pages are way too massive right now to be worth switching to the new UI. &#32; Headbomb {t · c · p · b} 17:53, 30 June 2019 (UTC)
 * Thanks for the fast responses. Izno: to repeat what I stated above: I did spend quite a bit of time trying to find which page contained the actual coding for the old URL, and I was unsuccessful. It's not obvious that discussion of Template: fixes should go on a talk page in the Help: namespace. My search obviously missed the archival talk page Help talk:Citation Style 1/Archive 56. Anyway, I'll continue in the non-closed thread. Boud (talk) 18:41, 30 June 2019 (UTC)

Update URL for bibcode tag
Please update the URL for bibcode=..., for example bibcode=1974AJ.....79..819H form http://adsabs.harvard.edu/abs/1974AJ.....79..819H to https://ui.adsabs.harvard.edu/abs/1974AJ.....79..819H. ADS Classic is now deprecated. It will be completely retired in October 2019. Read here: https://adsabs.github.io/blog/transition-reminder https://adsabs.github.io/blog/ave-atque-vale — Preceding unsigned comment added by Infin2694529 (talk • contribs) 19:30, 25 June 2019 (UTC)
 * I would personally hold on from that until the last moment we could. The old ADS site is much, muuuuch faster and does not require javascript. &#32; Headbomb {t · c · p · b} 20:18, 25 June 2019 (UTC)
 * Thank you for taking these technical issues in consideration, — Paleo Neonate  – 20:22, 25 June 2019 (UTC)
 * There's a detailed, archived thread here: Help_talk:Citation_Style_1/Archive_56. The most important response from Kelly from the ADS seems to me to be "in actuality we'll just be taking down the search pages and user accounts. Existing abstract page links will still work indefinitely, though they'll likely redirect to the equivalent pages in the new system at some point in the future."" Given the discussion there, I don't have a strong opinion about whether we should update Template:cite journal and Template:bibcode to the new URL (though switching from http:// to https:// would definitely be justified). Some arguments: In favour of updating: if a URL is going to become a redirect, making it direct will reduce a little bit of wasted cpu power and bandwidth. In favour of delaying: there's no big risk in waiting until the switch to a redirect occurs; delaying will reduce the amount of wasted javascript nonsense and need for uMatrix or similar blocking. I guess we should update the documentation where people expect to find it - probably a brief note at Template:Cite_journal/doc. Boud (talk) 18:57, 30 June 2019 (UTC)
 * Boud (talk) 19:35, 30 June 2019 (UTC)

Check for final character in several regular fields
For example, no author name ends with a  or   or. We could have a check for to make sure they don't end with such a comma, colon, or semi-colon, giving things like
 * lastn/firstn/authorn/collaboration

Likewise for a slew of parameters, which are basically every parameter except
 * url
 * article-url
 * chapter-url
 * contribution-url
 * Any other parameters (e.g. arxiv) which have their own, more-stringent, checks.

I also propose they are initially made as maintenance categories so they don't get thrown as big red errors to readers while kinks get worked out, and (possibly) corner cases identified.

The only parameters I see this as potentially problematic is chapter/title/quote (for technical documentation, like Reasons not to use & over &amp;amp;).

&#32; Headbomb {t · c · p · b} 21:19, 25 June 2019 (UTC)
 * I've cleaned up ~3000 such so far. Haven't seen an issue with the proposed logic above (for commas), although I did avoid touching chapter/title. I'll be tackling colons and semicolons next. &#32; Headbomb {t · c · p · b} 00:32, 27 June 2019 (UTC)


 * Dash (-) and its variants is another common trailing garbage character. Yes, please skip URLs. --  Green  C  01:00, 27 June 2019 (UTC)


 * Checking for these characters is fine, as they indicate a possible problem. But assuming that the problem is simply the presence of such characters, which problem is fixed by removing those characters, invites a bigger problem: overlooking that something might be missing. These characters all imply continuation. E.g., where Chapter 1: really should be Chapter 1: Introduction. Or A History of the World: is really A History of the World: the Last Five Years. The seemingly extraneous terminal character is actually a signal, or at least a clue, that something else is missing. Looking at this a minor cleanup of extraneous characters removes those clues. If they were to be the basis of a maintenance category there should be an instruction to first check the source to if the chapter/title/whatever needs to be completed. &diams; J. Johnson (JJ) (talk) 19:52, 27 June 2019 (UTC)
 * Well regardless, Chapter 1 is still correct and better Chapter 1:, likewise for A History of the World. But that's the sort of thing we'd find out through maintenance categories. This is also why chapter/title are also a bit different from pages, where a 32: is clearly an error. &#32; Headbomb {t · c · p · b} 20:04, 27 June 2019 (UTC)


 * No. "Chapter 1" is better only if that is the full name of the chapter. If the chapter is actually titled "Chapter 1: Introduction", then that is the correct – and thereby better – value. I grant you pagination, but with titles mindless lopping off a character because it is terminal, when really it should be intermedial, loses an indication that the title might be incomplete. &diams; J. Johnson (JJ) (talk) 23:32, 27 June 2019 (UTC)
 * See truncation and subtitle for why people might refer to Newton's work as The First Three Minutes rather than The First Three Minutes: A Modern View of the Origin of the Universe. Either way, the maintenance category would be appropriate. &#32; Headbomb {t · c · p · b} 00:49, 28 June 2019 (UTC)
 * Because I was curious about how this might be solved, I've hacked the sandbox:
 * This version of the code looks at every parameter value in a citation except for the parameters that convert to these internal meta-parameters:
 * If the code finds any one of the characters in the set  then it adds the article to.
 * —Trappist the monk (talk) 23:16, 28 June 2019 (UTC)
 * isn't there also a delimiter or seperator or something that should be added to the excepted parameters? &#32; Headbomb {t · c · p · b} 17:44, 29 June 2019 (UTC)
 * There used to be author-name-separator, author-separator, editor-name-separator,  editor-separator,  name-separator,  and separator.  Those all went away when we adopted mode.
 * —Trappist the monk (talk) 19:15, 29 June 2019 (UTC)
 * If the code finds any one of the characters in the set  then it adds the article to.
 * —Trappist the monk (talk) 23:16, 28 June 2019 (UTC)
 * isn't there also a delimiter or seperator or something that should be added to the excepted parameters? &#32; Headbomb {t · c · p · b} 17:44, 29 June 2019 (UTC)
 * There used to be author-name-separator, author-separator, editor-name-separator,  editor-separator,  name-separator,  and separator.  Those all went away when we adopted mode.
 * —Trappist the monk (talk) 19:15, 29 June 2019 (UTC)
 * —Trappist the monk (talk) 19:15, 29 June 2019 (UTC)

Also, I think it's worth thinking about expanding the set to other characters e.g., ... They don't all have to be done right away, but it's something to think about. It would be good to know what the explicit list of parameters that would have this check so we can see if there are corner cases (e.g. final - in author is an issue, but a final - in pages might not be (e.g. 33-), or have a customized list of checks, e.g. pseudocode finalcheck, = allowed {quote, url} . = disallowed {last, editor-last, year, date, volume, issue, pages, page, at, ...} & = disallowed {all} $ = allowed {chapter, title, quote, ...} &#32; Headbomb {t · c · p · b} 17:59, 29 June 2019 (UTC)
 * = allowed {chapter, title, quote, ...}

Commonly found trailing garbage in URLs includes: -- Green  C  23:56, 28 June 2019 (UTC)
 * [.,;:-l]$ (eg. example.com/index.htmll)
 * " $ (eg. example.com/index.htm"Title")
 * [']{2} $ (eg. example.com/index.htm Title )
 * [%][a-z0-9][a-z0-9]$ (exceptions: %22, %29, %5B, %7B) (eg. example.com/index.htm%7C)
 * The thing is, urls can end many of those and remain valid. Ending with a comma or a period for example, is perfectly fine, even if it will sometimes be wrong. &#32; Headbomb {t · c · p · b} 17:42, 29 June 2019 (UTC)

right, sometimes wrong: How often it is not-wrong for comparison is the question. Which is worse tracking or not. -- Green  C  05:11, 30 June 2019 (UTC)
 * Cameron Price has . With the trailing garbage character removed ("l"), the link works . A trailing "l" is a common occurrence on Wikipedia including cases like .pdfl
 * Carrie Mac has . With the trailing garbage character removed (".") the link works
 * Kilncadzow has . With the trailing garbage character removed (",") the link works
 * Given you'll have .html endings, or plenty of &value=something with a something that ends in l, I can't see that the false positives would not overwhelm the actual problems. &#32; Headbomb {t · c · p · b} 09:38, 30 June 2019 (UTC)
 * Just check for double l's or "htmll" or "pdfl" etc.. -- Green  C  15:18, 30 June 2019 (UTC)
 * That could be a separete check for invalid file extensions. Just a final character check isn't good enough for those. &#32; Headbomb {t · c · p · b} 17:54, 30 June 2019 (UTC)
 * It's a trailing garbage check. Limiting it to a single character is arbitrary. When you add something let me know in the mean time my bot keeps fixing problems. -- Green  C  06:09, 1 July 2019 (UTC)

How to handle a pseudonym
A book that appeared with the author given on the title page as "A Virginian", but you know his real name, how is this best handled? I wish there were a "pseudonym" field.

Please don't get off on how we know the real name. That's a different topic.

This was posted unsuccessfully at Teahouse. Thanks for any help. deisenbe (talk) 01:42, 5 July 2019 (UTC)
 * You cite the work as it appears in the copy you referenced. --Izno (talk) 02:30, 5 July 2019 (UTC)


 * You can also mention the author's actual identity in the article's prose (ideally with a citation confirming the identity), while leaving the citation to have however the author is listed in the source. As to how to include it, CMOS 17 has some guidelines which you might be interested in: If a work is attributed to an invented or descriptive name, and the author’s real name is not known, pseud. (roman, in brackets) may follow the name, especially if it might not be immediately clear to readers that the name is false (as in the first two examples below). (An initial The or A may be omitted. In a text citation, or in a shortened form in a note, pseud. is usually omitted.), A widely used pseudonym is generally treated as if it were the author’s real name., The real name, if of interest to readers, may follow the pseudonym in brackets., If the author’s real name is better known than the pseudonym, the real name should be used. Umimmak (talk) 04:31, 5 July 2019 (UTC)
 * Alas, CMOS is not cs1|2. Editor Izno is correct.  Cite the author as identified in the work that you consulted.  Extraneous text of any sort does not belong in cs1|2 template parameters.
 * —Trappist the monk (talk) 10:46, 5 July 2019 (UTC)
 * Which is not to say that notes cannot be left outside the citation-proper, whether inside the ref tag or in the article-proper. If the author or even his pseudonym is notable, one can also link the page instead of either of those two workarounds. --Izno (talk) 13:53, 5 July 2019 (UTC)
 * If the author has a separate article, you can also use an article-link parameter to link to the article. If you do that, the linked article should mention the pseudonym, to avoid confusing readers. —David Eppstein (talk) 20:53, 5 July 2019 (UTC)

Display problem with issue parameter
When  is used in cite news (e.g. to cite a daily newspaper), the number is displayed as 10, 000 (with a space), which goes against MOS:DIGITS. If  is used, template shows 10000 (without a comma), which is also against the Manual of Style. produces the same result, but at least it can be avoided, as 1000 is accepted by the MOS: Numbers with exactly four digits left of the decimal point may optionally be grouped (either 1,250  or  1250), provided that this is consistent within each article. SLBedit (talk) 01:15, 14 July 2019 (UTC)
 * Related discussion.
 * —Trappist the monk (talk) 03:01, 14 July 2019 (UTC)

Books with volumes and parts
Some books are differentiated from each other by volumes and parts. For example: How would I cite the last two? If I use the "issue" parameter in 'cite book' template it doesn't show: --Brianann MacAmhlaidh (talk) 22:50, 3 July 2019 (UTC)
 * "Smith (1900) History of Apples, Vol. 1"
 * "Smith (1901) History of Apples, Vol. 2, Part 1"
 * "Smith (1902) History of Apples, Vol. 2, Part 2".
 * There is no 'official' way to do what you want. Why?  Don't know.  Insufficient interest or need?  The metadata standard for books that  uses doesn't support part.  The metadata standard doesn't support issue or number for books either which is why  doesn't support those parameters.  One way to accomplish what you want is this:
 * If you are citing specific pages in that part, then pt. 1 pp. 75–83
 * —Trappist the monk (talk) 00:13, 4 July 2019 (UTC)
 * If you are citing specific pages in that part, then pt. 1 pp. 75–83
 * —Trappist the monk (talk) 00:13, 4 July 2019 (UTC)


 * I've had the same issue. What I end up doing is just with cs1, as in
 * It'd be great if allowed both volume and issue, but this work-around ends up producing the right results. Umimmak (talk)
 * Don't rely on that work-around remaining. It is a bug that should be fixed.
 * —Trappist the monk (talk) 11:50, 4 July 2019 (UTC)
 * Why would this be considered a bug that needs fixing? Umimmak (talk) 23:05, 4 July 2019 (UTC)
 * From the documentation:
 * "If the correct parameters are used, this template produces output identical to that of the Cite templates, such as Cite book and Cite web, ..." (The documentation goes on to talk about separators and automatic anchor generation.)
 * does not support issue so book citations rendered with should do the same.  The journal-style volume and issue format rendered by your example looks like a journal cite without an article title – a style that is allowed in journal citations where none.  As you can see, your example mimics that style but produces invalid metadata were it used to create journal citations without article title.
 * —Trappist the monk (talk) 00:07, 5 July 2019 (UTC)
 * One thing I've love to have is that books would display "Vol. " (or the full "Volume" or "vol." or whatever). I.e.
 * Sunada, T. (2013). "Generalities on Graphs". In Bloch, A.; Epstein, C.L.; Goriely, A.; Greengard, L. (eds.). Topological Crystallography. Surveys and Tutorials in the Applied Mathematical Sciences. Vol. 6, pp. 21–35. Tokyo: Springer. . ISBN 978-4-431-54177-6.
 * instead of
 * which is completely unclear. And even group Vol. 6, pp. 21–35. together for sanity.&#32; Headbomb {t · c · p · b} 02:18, 4 July 2019 (UTC)
 * I am fairly certain the bolding in cite book is new, and is basically cite journal leaking its styling... but that may just be my bad memory. --Izno (talk) 03:14, 4 July 2019 (UTC)
 * , even in olden days, bolded volume:
 * which is completely unclear. And even group Vol. 6, pp. 21–35. together for sanity.&#32; Headbomb {t · c · p · b} 02:18, 4 July 2019 (UTC)
 * I am fairly certain the bolding in cite book is new, and is basically cite journal leaking its styling... but that may just be my bad memory. --Izno (talk) 03:14, 4 July 2019 (UTC)
 * , even in olden days, bolded volume:


 * —Trappist the monk (talk) 11:50, 4 July 2019 (UTC)

A related issue is that a book may have multiple volumes for its particular title, and also have a different numbering as a volume within a book series. What I've usually resorted to in such cases is to put, e.g. "Vol. I" into the title parameter, and use the volume parameter for the book series volume number. —David Eppstein (talk) 19:01, 4 July 2019 (UTC)

As I noted earlier in this discussion, is broken. I've tweaked the sandbox to fix that:

book and encyclopedia cites do not show issue: periodical cites (journal, magazine, news) show issue: web cites (website) show neither volume nor issue: script-&lt;periodical> except script-website show both volume and issue: with script-website shows neither volume nor issue: cs1 templates to show that I haven't broken anything: with script-&lt;periodical>: While making these fixes, I also noticed that  without url does not show the missing url error message. Fixed that:

—Trappist the monk (talk) 16:44, 5 July 2019 (UTC)

I've never liked the way "volume" appears for books. The template uses the form appropriate for journals. Would it be possible for it to render as "Vol. XXX" like suggested above? Hawkeye7  (discuss)  01:36, 15 July 2019 (UTC)

Fill in automatically
When adding an archive link, the  has to be entered manually, even if the   already contains it: https://web.archive.org/web/20020120142510/http://example.com/ Many other archives besides Wayback include the date (also the source URL) in the snapshot URL as well.

Couldn't this be automated? That would be more convenient, and less error-prone. Toxide (talk) 10:57, 17 July 2019 (UTC)
 * This was also cross-posted at . One conversation in one place, I have answered here.
 * Related discussion:
 * I can imagine that and the cs1|2 modules might share some bit of code that would extract an archive date from archive-url –  already does this.  I have not determined whether or not that functionality is easily shared without substantial changes to either or both of cs1|2 and.
 * In the best of all possible worlds, it would seem to me that if we are to adopt this, archive-date should go away entirely but that would mean that anything put in archive-url must have an encoded archive date. At present, there is no requirement that archive-url hold a url of an archival service (though best-practices would suggest that non-archival-service urls are suspect).
 * —Trappist the monk (talk) 14:35, 17 July 2019 (UTC)
 * In the best of all possible worlds, it would seem to me that if we are to adopt this, archive-date should go away entirely but that would mean that anything put in archive-url must have an encoded archive date. At present, there is no requirement that archive-url hold a url of an archival service (though best-practices would suggest that non-archival-service urls are suspect).
 * —Trappist the monk (talk) 14:35, 17 July 2019 (UTC)
 * —Trappist the monk (talk) 14:35, 17 July 2019 (UTC)


 * Not all archive services use 14-digit timestamps, or don't use them by requirements. WP:WEBARCHIVES lists the archive providers in use on Enwiki (there are others in other languages not listed) though new archive providers do appear may not be listed here. WP:WEBARCHIVES shows the URL formats we are dealing with and any software would be expected to parse. As you can see, not all use 14-digit timestamps. -- Green  C  15:14, 17 July 2019 (UTC)
 * It is possible for the software to compare the date in archive-date with the 14-digit timestamp, if it exists, and generate a red tracking message about a mismatch. already does this and CS1|2 might also. My bot WaybackMedic is able to fix these, and on occasion runs through the  tracking category to clear it.  --  Green  C  15:19, 17 July 2019 (UTC)

Author affiliation for cite journal? (As in the university an author works at), and also adding editors of books with chapters written by different people?
1. Has anyone considered adding author affiliation as a feature of cite journal? For example in Ottoman_constitution_of_1876 I wanted to state the university affiliation of a journal article author, but that would mean putting parentheses in the author field. Is there a way for the template to display affiliation in this way without having parentheses in the author field?

2. I haven't used citation templates to cite works where a book contains multiple essays written by different people, as you have the author of the essay itself and then the editors of the book in question. Is this functionality there in the template? If not, would it be OK to add it?

Thanks, WhisperToMe (talk) 21:32, 17 July 2019 (UTC)
 * An affiliation is not part of an author's name so doing as you did at :
 * corrupts the author metadata (not only the parentheses but also the wikimarkup).
 * What is really missing from that citation template is the journal. With the next release of the cs1|2 module suite,  will require journal
 * Edited books with multiple section / chapter / entry authors is supported:
 * —Trappist the monk (talk) 21:53, 17 July 2019 (UTC)
 * Thanks for the info! In that case it would be nice to add an "affiliation" function so that can be noted after each author. As for the editor function, I'll try that. I'll see if multiple editors may be supported WhisperToMe (talk) 23:16, 17 July 2019 (UTC)
 * I was able to get an instance working at History of the Central Americans in Houston! I wonder how to get both ISBNs in, though, as I looked at the documentation and couldn't see how to enter ISBN-10 and ISBN-13 separately WhisperToMe (talk) 23:22, 17 July 2019 (UTC)
 * I've never seen a citation style where the institutional affiliation of each author gets included; that seems like unnecessary clutter. editorN-first and editorN-first can support different numbers of N to allow multiple editors. I also don't see the reason to include an ISBN-10 when you have access to an ISBN-13. Just cite the ISBN-13. Umimmak (talk) 23:51, 17 July 2019 (UTC)
 * In many academic journal articles it's common to name the university the author works at, and to do so prominently (in the lead of the article and/or with a footnote attached to the name). Based on the prominence there, I thought that in academic journal citations it's important to indicate the author's institution in parentheses. WhisperToMe (talk) 00:11, 18 July 2019 (UTC)
 * It's common to do so in the article itself, but I have never seen it in a citation – our citations have enough overkill already. Triptothecottage (talk) 00:14, 18 July 2019 (UTC)
 * But those don't ever appear in . Journal articles often also include the authors' email addresses but it would be overkill to have that in a citation to that paper. If the authors' institutional affiliation is essential, then you can mention it in the article prose -- not in the citation. Umimmak (talk) 00:26, 18 July 2019 (UTC)
 * I agree with others, it is non-standard in academic writing to include the author's affiliation in the citation. If it's important, it's important enough to mention in the text instead of in the citation, like this
 * "John Doe, director of the Institute of Dust Bunny Studies, stated that....34"
 * In many academic journal articles it's common to name the university the author works at, and to do so prominently (in the lead of the article and/or with a footnote attached to the name). Based on the prominence there, I thought that in academic journal citations it's important to indicate the author's institution in parentheses. WhisperToMe (talk) 00:11, 18 July 2019 (UTC)
 * It's common to do so in the article itself, but I have never seen it in a citation – our citations have enough overkill already. Triptothecottage (talk) 00:14, 18 July 2019 (UTC)
 * But those don't ever appear in . Journal articles often also include the authors' email addresses but it would be overkill to have that in a citation to that paper. If the authors' institutional affiliation is essential, then you can mention it in the article prose -- not in the citation. Umimmak (talk) 00:26, 18 July 2019 (UTC)
 * I agree with others, it is non-standard in academic writing to include the author's affiliation in the citation. If it's important, it's important enough to mention in the text instead of in the citation, like this
 * "John Doe, director of the Institute of Dust Bunny Studies, stated that....34"


 * Jc3s5h (talk) 10:29, 18 July 2019 (UTC)
 * Fair enough... I'll just mention it in the prose. WhisperToMe (talk) 11:37, 18 July 2019 (UTC)

Another thing I'd like to explore: Is there a function where a URL can be given for a particular page in a book and that makes the page number clickable? The reason why I want this is that I want the title of the book to serve as a Wikipedia link to an article about the book itself (I often intentionally write Wikipedia articles about books used as sources so readers can learn about the background of the book), while the page number would be the place where one can click for the source content. WhisperToMe (talk) 11:37, 18 July 2019 (UTC)
 * —Trappist the monk (talk) 12:03, 18 July 2019 (UTC)
 * Thank you! WhisperToMe (talk) 12:21, 18 July 2019 (UTC)
 * —Trappist the monk (talk) 12:03, 18 July 2019 (UTC)
 * Thank you! WhisperToMe (talk) 12:21, 18 July 2019 (UTC)

access icon parameter value bug
This bug has been in the code since we added support for the access icons. Surprising that it's not been noticed until now. Fixed in the sandbox:

—Trappist the monk (talk) 22:28, 17 July 2019 (UTC)
 * Thank you for this fix! − Pintoch (talk) 12:12, 19 July 2019 (UTC)

Why no doi-access=subscription ?
Why is the option doi-access=subscription not allowed when in fact most DOI lead to journal websites that require subscription or access via university libraries? --bender235 (talk) 15:38, 15 July 2019 (UTC)
 * The norm in access requirements is not indicated. However, I believe a registration option should be allowed for subscription-normal content identifiers such as doi or jstor. I have encountered registration-required access situations for both. 72.43.99.130 (talk) 15:46, 15 July 2019 (UTC)
 * There is a url-access=subscription, which is why I was asking. --bender235 (talk) 15:49, 15 July 2019 (UTC)
 * Because, for named identifiers, the linked source usually lies behind some sort of registration barrier or paywall. When that isn't the case, that is when (for doi as an example) free should be used.  For url, the norm is that the linked source is usually free-to-read.  When that isn't the case, that is when subscription, registration, or limited should be used.
 * This is explained in the template documentation: . Is the documentation insufficient?
 * —Trappist the monk (talk) 16:13, 15 July 2019 (UTC)
 * Maybe it would be good to accept the value instead of reject it and then display nothing, as well as a maintenance message. --Izno (talk) 17:50, 15 July 2019 (UTC)
 * meh. Then no one learns anything and it becomes busy work for some bot or dedicated gnome to fix.  And we'll start getting: 'I added subscription but it isn't showing the red lock.  How come?'
 * —Trappist the monk (talk) 20:01, 15 July 2019 (UTC)
 * As mentioned above, jstor should be added for content that requires registration only. Without this clarification, readers are not offered an existing path to non-paid verification. 65.88.88.91 (talk) 18:24, 15 July 2019 (UTC)
 * That will just add complication and confusion. The rule is as simple as I think we can make it: sources linked by the named identifiers are behind some sort of restrictive barrier unless marked otherwise with free.
 * —Trappist the monk (talk) 20:01, 15 July 2019 (UTC)
 * I don't understand who would be confused by this? There is already a similar array of access options for url and I cannot remember anybody complaining about it. The same array of options (with allowances for the access norm) should be extended to all content identifiers. The specific case has the potential to make verifiability easier/more accessible. I would think this would make it a no-brainer. 65.88.88.91 (talk) 20:39, 15 July 2019 (UTC)
 * Oh, so for DOI "subscription" is the default, and "free" would be the exceptional case. I see. But shouldn't the template then display the little red lock Lock-red-alt-2.svg by default, too? --bender235 (talk) 19:31, 15 July 2019 (UTC)
 * Not quite. For most identifiers, the default is: not-free-to-read.  The default state says nothing about what kind of restrictions the publisher has placed on the source.  We do not highlight the norm and, because there are subscription, registration, and limited access options, it is not possible for the templates to apply an appropriate lock icon without being told what that lock icon should be.
 * —Trappist the monk (talk) 20:01, 15 July 2019 (UTC)
 * we had this exact debate when we implemented this - I was of the same opinion as you, but we could only get the change to pass in this form. − Pintoch (talk) 12:14, 19 July 2019 (UTC)
 * we had this exact debate when we implemented this - I was of the same opinion as you, but we could only get the change to pass in this form. − Pintoch (talk) 12:14, 19 July 2019 (UTC)

On requiring the journal parameter
I read above that:
 * With the next release of the cs1|2 module suite, will require journal

Are we sure about this? There are thousands of journals which use DataCite DOIs, but the funny thing (as someone highlighted to me at some recent conferences) is that the DataCite schema doesn't have a field for the journal name. This might be dismissed as irrelevant but is a sign of what is considered important. Nemo 13:27, 19 July 2019 (UTC)
 * I see no reason not to require journal when using – that is, after all, the purpose of the template.
 * Shouldn't the data that support a journal article be cited from that journal article and not from a Wikipedia article?
 * Aren't the data that the author(s) used in their research and from which they drew their conclusions, a primary source that is not subject to the normal editorial processes of publication and review? (WP:PRIMARY)
 * If it is determined that citing research data is an appropriate thing to do, then shouldn't cs1|2 have a template specific to that requirement so that editors don't misuse other templates?
 * —Trappist the monk (talk) 14:42, 19 July 2019 (UTC)
 * Primary sources can be used in Wikipedia with care. One might imagine, for example, a definition that needs to be precisely stated in a Wikipedia article, but the term is undefined, or vaguely defined in a journal article; the precise definition needs to be retrieved from the data. Or perhaps two different journal articles appear to rely on the same data, and citations to the data for each article are needed to confirm this.
 * The problem with a special template that doesn't exist is that an editor can't be expected to wait for one of the few people who understand the monster that the citation templates have become to provide a special template. The editor would hand-code a citation. The editor would even be justified in rewriting every citation in the article to some other style that doesn't require templates, to avoid the suppression of acceptable sources by inadequate software. Jc3s5h (talk) 15:07, 19 July 2019 (UTC)
 * I don't think that I said that primary sources could not be used. I do question the validity of citing research data here, at Wikipedia, because raw data are so often subject to interpretation which editors here should not be doing.  I accept that there may be the occasional need to cite research data.  If we are going to support that, we should provide a template specifically tailored to that task.
 * Yeah, such a template doesn't exist. So what?  If it is needed it will be created and yeah, that creation won't be instantaneous.  Again, so what?
 * —Trappist the monk (talk) 18:55, 19 July 2019 (UTC)
 * There is a reason, cite paper and cite document redirect to cite journal. &#32; Headbomb {t · c · p · b} 16:07, 19 July 2019 (UTC)
 * I spent a few minutes looking at a few dozen and  transclusions.  From that small sample, it seems to me that most could be (should be) rewritten to use a more appropriate template; obvious choices for the citations that I inspected were:, , , , , , .  Template redirects do not act any differently from the template that they alias; editors should not expect different action simply because of a redirect's name.  Perhaps this exercise will show that we need separate  and  templates.
 * —Trappist the monk (talk) 18:55, 19 July 2019 (UTC)
 * Perhaps they should be rewritten before you break them all by making journal mandatory? —David Eppstein (talk) 22:08, 19 July 2019 (UTC)
 * If the objective is to require that "cite journal" is only used for documents which belong to some kind of "journal", also known as "papers", I suspect that requiring to provide a journal name is not the easiest or most efficient way to do it. Nemo 17:12, 19 July 2019 (UTC)
 * is a periodical template – always has been. Individual papers and documents are not periodicals.  We already have some templates that are suitable for citing preprints of individual papers so it isn't much of a stretch to imagine that we should also have templates specifically for citing individual papers and documents – the  redirect might be usurped for that purpose and the  redirect pointed at  instead of.
 * You suspect that requiring to provide a journal name is not the easiest or most efficient way to ensure that journal cites name the journal. State the better method; don't just leave us hanging ...
 * —Trappist the monk (talk) 18:55, 19 July 2019 (UTC)
 * I'd be more than fine having a fork of cite paper/cite document as a distinct template from cite journal if we're going to make journal mandatory. Likely first as a maintenance category to see how bad things are have a preliminary round of cleanup through citation bot and the like. &#32; Headbomb {t · c · p · b} 20:42, 19 July 2019 (UTC)
 * The better method is something both understandable to users and easy to verify. I would personally prefer a superset of the publications eligible for plan S, namely one of their requirements which is easy to use: «Use of persistent identifiers (PIDs) for scholarly publications (with versioning, for example, in case of revisions), such as DOI (preferable), URN, or Handle».
 * On the other hand, if the objective is to prevent usage of this template by articles or depending templates for unforeseen citations which end up creating undue style demands or other hurdles for the maintainer(s) of this template, I don't have an easy answer. I wish we could fix the (technical? social?) problem with those usages rather than force their hand by breaking them. Nemo 09:34, 20 July 2019 (UTC)
 * Plan S is fine and all, but that retroactively make some self-published 1928 paper from Johann Gambolputty all of a sudden become published and have a DOI. &#32; Headbomb {t · c · p · b} 10:08, 20 July 2019 (UTC)
 * Surely we don't expect a wikitext condition to be able to take care of such undesired events though. Nemo 10:41, 20 July 2019 (UTC)
 * Sure. With cite paper. Not cite journal. &#32; Headbomb {t · c · p · b} 11:02, 20 July 2019 (UTC)
 * I think that you are describing a solution to a problem that isn't the journal-cite-without-journal-title problem. I do not think that there should be any requirement for persistent identifiers in  because there likely to be sources that can use  that lived and died before persistent identifiers were invented.  Nor do I think that a persistent-identifier-requirement is a solution to your research data concern.
 * An actual template to replace the redirect to  would handle individually published papers that do not belong to some kind of "journal".  This template can use but must not require persistent identifiers, does not allow journal or aliases, gets the metadata right.   should require publisher because WP:V.  These same might apply to a  template were we to decide that citing research data is something that cs1|2 should support.
 * —Trappist the monk (talk) 11:13, 20 July 2019 (UTC)
 * Surely we don't expect a wikitext condition to be able to take care of such undesired events though. Nemo 10:41, 20 July 2019 (UTC)
 * Sure. With cite paper. Not cite journal. &#32; Headbomb {t · c · p · b} 11:02, 20 July 2019 (UTC)
 * I think that you are describing a solution to a problem that isn't the journal-cite-without-journal-title problem. I do not think that there should be any requirement for persistent identifiers in  because there likely to be sources that can use  that lived and died before persistent identifiers were invented.  Nor do I think that a persistent-identifier-requirement is a solution to your research data concern.
 * An actual template to replace the redirect to  would handle individually published papers that do not belong to some kind of "journal".  This template can use but must not require persistent identifiers, does not allow journal or aliases, gets the metadata right.   should require publisher because WP:V.  These same might apply to a  template were we to decide that citing research data is something that cs1|2 should support.
 * —Trappist the monk (talk) 11:13, 20 July 2019 (UTC)
 * An actual template to replace the redirect to  would handle individually published papers that do not belong to some kind of "journal".  This template can use but must not require persistent identifiers, does not allow journal or aliases, gets the metadata right.   should require publisher because WP:V.  These same might apply to a  template were we to decide that citing research data is something that cs1|2 should support.
 * —Trappist the monk (talk) 11:13, 20 July 2019 (UTC)

Break Category:Pages with DOIs inactive as of 2019 into months
3000 pages makes it very hard to check what's backlogged vs what's newly broken. Breaking down by months would make it easier to manage. &#32; Headbomb {t · c · p · b} 07:19, 20 July 2019 (UTC)
 * I've tweaked Module:Citation/CS1/Identifiers/sandbox so that it adds a sortkey to the category when it can decode a month value from doi-broken
 * To prove that it is working, copy this:
 * into Special:ExpandTemplates (the Input wikitext box) to see the category with sortkey (cs1|2 does not add categories when in the Help talk namespace). Change the date, give it invalid date, etc.
 * When not given a month, does not add a sortkey. We might want to change that so that month-less articles don't sort by article-name first-letter.  Might also want to change the sort key from month-name to month number.
 * Another option is to create subcategories:
 * —Trappist the monk (talk) 13:18, 20 July 2019 (UTC)
 * I would much prefer monthly categories. A sortkey, numerical or not, would not let you easily separate items by date (J = January, June, July; 1 = January, October, November, December). &#32; Headbomb {t · c · p · b} 17:19, 20 July 2019 (UTC)
 * Ok,
 * —Trappist the monk (talk) 10:59, 21 July 2019 (UTC)
 * —Trappist the monk (talk) 13:18, 20 July 2019 (UTC)
 * I would much prefer monthly categories. A sortkey, numerical or not, would not let you easily separate items by date (J = January, June, July; 1 = January, October, November, December). &#32; Headbomb {t · c · p · b} 17:19, 20 July 2019 (UTC)
 * Ok,
 * —Trappist the monk (talk) 10:59, 21 July 2019 (UTC)

cite ssrn
Harking back to this discussion, I found this:

and that lead me to create to become  after the next module update:

Like, , and , this template accepts only a limited subset of the cs1|2 parameters.

Keep? Discard?

Needs documentation which I leave to someone else.

I will note that still needs documentation ...

—Trappist the monk (talk) 23:56, 18 July 2019 (UTC)


 * The solution, I feel, is to create a cite preprint and ultimately redirect cite arxiv/biorxiv/citeseerx/ssrn there, or have cite arxiv be a wrapper of  or something. But that might be more complex than individual templates.&#32; Headbomb {t · c · p · b} 00:57, 19 July 2019 (UTC)
 * Yeah, maybe. Not clear to me how such a template would handle metadata if it were presented with more than one identifier (specifically  ).  That suggests that only one of arxiv (or eprint), biorxiv, citeseerx, or ssrn would be allowed in any one instance of .  That means that the preprint parameters would all be pseudo-aliases of each other for the purpose of duplication detection (and error messaging) but as individual parameters for the purpose of rendering the proper link.
 * —Trappist the monk (talk) 11:53, 19 July 2019 (UTC)

|url=?
I just ran across this:

This template makes me wonder if supporting url in and the other preprint templates is the correct thing. For this particular example, I would say no. In this case, the paper was presented at the Twenty-eighth Conference on Neural Information Processing Systems (NIPS 2014 video of the talk linked from the conference schedule). So, this template is one that is better rewritten to use (which I will do).

Is there any reason that the identifier-based preprint templates need url?

—Trappist the monk (talk) 14:25, 21 July 2019 (UTC)

Name order
As it is, the cite book template puts the surname before the first name (e.g. "Smith, John"). I'd like someone to add a option that will force the template to order it first then last (ie "John Smith"). The only way to do this currently is to use the author= parameter, but then this causes problems with the ref= parameter and other things. Could somebody do this for me? Something like a parameter first-last-name=true. Kurzon (talk) 17:37, 15 June 2019 (UTC)
 * I support some version of this. Specifically, with the af parameter, as I proposed it in 2018. &#32; Headbomb {t · c · p · b} 18:50, 15 June 2019 (UTC)
 * Combined with automatic global formatting like have for dates, especially. &#32; Headbomb {t · c · p · b} 15:59, 18 June 2019 (UTC)
 * Why?
 * First-last ordering of the lead author's name in a citation is considered acceptable when the citations occur as isolated driplets across the foot of various pages because there are usually only one or two, and there is no question of ordering. On the other hand, when full citations are collected into lists it is standard to collate them in some kind of order, which is typically alphabetically by the first author's surname (family name). As most Western cultures put this name last, which is inconvenient for the primary sort-key, it is standard practice to invert the name into "last, first" order. I do not know of any reason why this should not be done for the lead author.
 * Variation on this point occurs regarding co-authors. That is an arguable point, and it seems to me it has been argued before. If it is going to be raised again I would expect a review of previous related discussions. &diams; J. Johnson (JJ) (talk) 21:10, 17 June 2019 (UTC)
 * Well, for one thing you have Asian names where the surname comes first. Previously, people just told me to use author. Kurzon (talk) 11:54, 18 June 2019 (UTC)
 * Another consideration is how certain style guides handle things. In The Chicago Manual of Style system of footnotes/endnotes, the name order is not inverted. CMOS does invert name order in a bibliography for the first author only. For those more familiar with CMOS's methods, ours may seem strange.  Imzadi 1979  →   12:42, 18 June 2019 (UTC)
 * Why? Because WP:CITEVAR mostly, and because this would allow us to have "John Smith (1903) Book of Stuff ..." types of citation with correct last/first name metadata. But also because it would let us easily support a plethora of styles, from CMOS, Bluebook, Vancouver, and many others while also ensuring full consistency and error checking on author format across the whole article by slapping the equivalent of use dmy dates on the article, like we do for dates. You could slap something like use Vancouver names, or or similar, and not have to micromanage and review every citation after bots, tools, and editors get involved. &#32; Headbomb {t · c · p · b} 16:28, 19 June 2019 (UTC)
 * I second this. The first-last name order is more readable, in particular if commas are part of a name or commas are used as name separators (as it happens in some style guides). This format is therefore used in many areas outside WP (and even in WP when not using cs1 citation templates). The last-first order has advantages as well, and I think it should remain the default. However, since the usage of the citation template framework is preferable over "free-style" citations for many reasons, it is important for the framework to support all major display variants, because otherwise some people will simply not use the templates.
 * I would therefore applaud the addition of an af= parameter and global templates like "Use lf/fl names", and in the long run hopefully the possibility to set the preferred format in the user preferences overriding such Use names or Use dates templates.
 * Although I consider the usage of abbreviated names as an anachronism being difficult to read and often causing confusion (and our MOS advises against the usage of abbreviations where possible), I would even support if an af= parameter would optionally support the automatic truncation of given names, because there might be a few areas where it's actually useful (for as long as this never becomes a default). This may also help to improve the quality of our references, because people could always specify the unabbreviated names, even if only abbreviated names were to be displayed in a specific scenario.
 * --Matthiaspaul (talk) 10:17, 23 June 2019 (UTC)
 * It would also let us get rid of the awfully specific vanc in favour of something scalable to other styles. &#32; Headbomb {t · c · p · b} 21:48, 25 June 2019 (UTC)
 * Please pay closer attention: citations in a note at the foot of a page – i.e., footnotes – are exactly where I said that author names are not inverted. In any kind of list, such as a bibliography, it is preferred to sort the entries, which is most commonly alphabetically by authors.
 * Kurzon: yes, with Asian style (also Hungarian) the family name (surname) comes first. If there is an possibility of confusion just use surname, which is a synonym for last. As long as we invert "last" name everything works out. If we don't invert (perhaps for co-authors), well, that would be ambiguous. The problem with using author or coauthor for this is we have no indication of which order the names are in. &diams; J. Johnson (JJ) (talk) 23:10, 18 June 2019 (UTC)

Da; there is no need for any kind of inversion. [That came in with my edit, but it's not my comment. &diams; J. Johnson (JJ) (talk) 20:31, 19 June 2019 (UTC)]

Ideally, at least for lists of citations in alphabetical order, a person from a culture where the given name is written first in running text would be "Washington, George" while a person from a culture where the surname is written first in written first would be "Mao Zedong". To achieve this ideal, it would be necessary to individually mark each name to show which convention applies, or at least, mark those that don't follow the default for an English-language publication, "Washington, George". Jc3s5h (talk) 17:38, 19 June 2019 (UTC)


 * That is exactly what the comma does: it indicates that a cultural-specific "normal" ordering has been inverted to put the indexed term first. A question that has been raised before (here, but also outside of WP) is whether it is proper to have a comma in "Mao, Zedong", which might imply that inversion was done. I think a better view to take is that the comma marks the index term, and (from the pov of citation) we don't really care whether inversion was necessary. &diams; J. Johnson (JJ) (talk) 20:36, 19 June 2019 (UTC)
 * It could also be a simple delimiter, indication delineation Mao, Tse Tung vs Mao Tse, Tung, rather than inversion. &#32; Headbomb {t · c · p · b} 21:11, 19 June 2019 (UTC)
 * Huh? The surname is this example is "Mao", not "Mao Tse". Why would "Mao Tse" get delimited from "Tung"? &diams; J. Johnson (JJ) (talk) 23:54, 19 June 2019 (UTC)
 * That's my point. In Surname, Given name the comma makes it clear where the delimitation is. It doesn't have to indicate an inversion. &#32; Headbomb {t · c · p · b} 00:06, 20 June 2019 (UTC)
 * Aren't we saying the same thing here? &diams; J. Johnson (JJ) (talk) 19:34, 20 June 2019 (UTC)
 * You're saying the comma indicates inversion. I'm saying it might simply indicate delineation. &#32; Headbomb {t · c · p · b} 19:43, 20 June 2019 (UTC)
 * Perhaps I was not entirely clear. What I meant above is that a comma can indicate that a Western-style first-last ordering has been inverted. But regardless of whether inversion was necessary to get the surname first, the comma marks the index term. Or to use your term, delimits it. Same thing, right? &diams; J. Johnson (JJ) (talk) 20:54, 26 June 2019 (UTC)

I read the rules and Wikipedia does not have a fixed citation style. The citation templates are optional, not mandatory. I could ignore them to do what I want. Kurzon (talk) 07:49, 20 June 2019 (UTC)
 * You could, but it's also why it's important that CS1 templates can accommodate small variations like that. For now, you can use author instead off last/first. &#32; Headbomb {t · c · p · b} 15:26, 20 June 2019 (UTC)


 * Templates are strongly encouraged, as otherwise the identification of sources and locating them can get murky, and verifiability (our key principle) is impaired. On the the other hand, using authors instead of first/last is deprecable. Yes, the documentation suggests using it, but that was never vetted, and one of these days (soon?) ought to be re-visited.
 * Getting back to Kurzon's initial request: I think that is a "small variation" we ought not to accommodate. I don't believe that is common practice here, and in bibliographies "last, first" is practically required. &diams; J. Johnson (JJ) (talk) 19:50, 20 June 2019 (UTC)
 * It's a variation that most definitely ought to be accommodated. The cost of not doing so is greater inconsistency amongst articles, and a greater refusal to adopt citation templates, and poorer metadata because authorn has to be used over lastn/firstn. &#32; Headbomb {t · c · p · b} 20:48, 20 June 2019 (UTC)
 * Misuse of author (and I have seen plenty of that) corrupts the metadata. But where has anyone rejected use of templates because they wanted to collate by personal name? This warrants a separate, deeper discussion. &diams; J. Johnson (JJ) (talk) 21:00, 26 June 2019 (UTC)

WP:CITESTYLE says there is no hard rule for name order. Kurzon (talk) 09:57, 21 June 2019 (UTC)
 * I strongly disagree that the use of author or authors could or should be deprecated. These parameters are necessary, to handle institutional authors (where the author of record is a committee or some such), and probably also necessary in some unusual circumstances to handle authors who are people but who do not have names that fit the first/last paradigm (the obligatory link). Perhaps they could be deprecated for instances where the author is a conventionally-named person or group of people but I don't see how that could be enforced. —David Eppstein (talk) 22:13, 19 July 2019 (UTC)
 * Authors is deprecated by maintenance message at this time, in case you were not aware. Author is supported and I expect it to be supported until we identify some alternative like org-author for organizational authors. --Izno (talk) 18:42, 21 July 2019 (UTC)
 * It's not just organizational authors. As I thought I already clearly stated. An example: I recently added a source whose author is credited only as "Jacob". I believe it to be a first name, but the template does not allow first= without last=. The only reasonable choice is to set author=Jacob. —David Eppstein (talk) 19:19, 21 July 2019 (UTC)
 * Or Riazuddin/Fayyazuddin/Plato/etc... &#32; Headbomb {t · c · p · b} 19:44, 21 July 2019 (UTC)
 * Nothing wrong with that. It's authors (plural) that is discouraged (it isn't really deprecated though I think that sometime in future it should be).  This parameter is discouraged because of its free-form nature.  It allows any number of names, usually human, but because human names are, per your obligatory link, so damn confounding, the module does not attempt to add the content of authors (plural) to the citation's metadata.
 * —Trappist the monk (talk) 19:53, 21 July 2019 (UTC)
 * Not true. What WP:STYLE says is that "Wikipedia does not have a single house style". It also says (here): "The full citations are listed in alphabetical order, according to the authors' surnames, at the end of the article in a "References" section." Also (here, underlining added): "General references are usually listed at the end of the article in a "References" section, and are usually sorted by the last name of the author or the editor." As to alternative sorting, I am not aware of any instances, in or out of WP, of sorting by first (personal) names.
 * Not true. What WP:STYLE says is that "Wikipedia does not have a single house style". It also says (here): "The full citations are listed in alphabetical order, according to the authors' surnames, at the end of the article in a "References" section." Also (here, underlining added): "General references are usually listed at the end of the article in a "References" section, and are usually sorted by the last name of the author or the editor." As to alternative sorting, I am not aware of any instances, in or out of WP, of sorting by first (personal) names.


 * But don't forget what I said before: not inverting an author's name is typically done only in isolated citations, such as found in foot notes (footnotes), at the foot of a printed page, where there are only a couple of citations, and no need to sort them. &diams; J. Johnson (JJ) (talk) 21:10, 26 June 2019 (UTC)

New OAbot BRFA
Editors interested in citation templates might want to chime in: Bots/Requests for approval/OAbot 3. − Pintoch (talk) 15:31, 25 July 2019 (UTC)

Chapter trans title
I suggest enabling this because it is intuitive if one has url and archive-url to have chapter-trans-title beside trans-title for title and chapter params respectivelly. --Obsuser (talk) 16:44, 26 July 2019 (UTC)
 * trans-chapter already exists:
 * —Trappist the monk (talk) 17:04, 26 July 2019 (UTC)
 * I think I could not find it in documentation. If not there should be added, if already there everything resolved. Maybe only to add alternative name if convenient. --Obsuser (talk) 12:14, 27 July 2019 (UTC)
 * trans-chapter is documented and has been for a long time; see.
 * —Trappist the monk (talk) 12:50, 27 July 2019 (UTC)
 * trans-chapter is documented and has been for a long time; see.
 * —Trappist the monk (talk) 12:50, 27 July 2019 (UTC)

Question about |id=
I'm working on a draft that is mainly cited to EBSCOHost databases because I can. For most sources, the URL I'm using is just the direct link to the EBSCOHost data from my institution. However, for one I have a better (more accessible) link and decided to have mercy on our poor readers and included it. However, I am still going to include an EBSCHost link because I want to and because I can. I just don't know how to mark the link as I would subscription. Was the exclusion of a id-access intentional? It would seem that portions of the module code use it as a keyword (Module:Citation/CS1/Configuration defines it, and it is references in Module:Citation/CS1/Identifiers), but that essentially refers to all the doi-access-like parameters if I understand things right. Am I supposed to use a seperate template for this? Help would be nice here. This page isn't in my watchlist, so a ping is appreciated. BTW, the real reason I use EBSCOHost is because that is just what I have through school. Since I'm already using it, I want as many of my citations to be as uniform as possible here. Regards, &#8211;  MJL &thinsp;‐Talk‐☖ 22:43, 27 July 2019 (UTC)
 * id-access is not the same as   The former is a non-existent template parameter while the latter is Lua table key.
 * Because id can hold anything it is not constrained to be an external link. It is just some sort of free-form something tacked on at the end of the named-identifiers list.  I don't recall discussion about id during the access icon discussions.  I suppose that we could add support for free IFF [&lt;url> label]; only   because id is an identifier parameter.
 * —Trappist the monk (talk) 00:50, 28 July 2019 (UTC)
 * —Trappist the monk (talk) 00:50, 28 July 2019 (UTC)

URL error
The following cite generates a URL error. After reading the "help" page, I think it is caused by the "x". But I did not see a way to fix it:
 * —User-duck (talk) 18:43, 26 July 2019 (UTC)
 * In the sandbox added 'company' to the list of hostnames that allow single-letter second-level domain names.
 * —User-duck (talk) 18:43, 26 July 2019 (UTC)
 * In the sandbox added 'company' to the list of hostnames that allow single-letter second-level domain names.


 * —Trappist the monk (talk) 19:28, 26 July 2019 (UTC)
 * Thank you
 * —User-duck (talk) 15:16, 27 July 2019 (UTC)
 * If I may nitpick, "company" is not a hostname but a dns suffix, signifying in this case a generic top-level domain. 172.254.255.250 (talk) 19:34, 28 July 2019 (UTC)

Sense number
Is there any way to cite a specific sense of an entry? I'm hoping for something like how the Wiktionary template (wikt:Template:R:OED_Online) has a `pos` (part of speech) parameter that allows you to cite a specific sense in a dictionary entry.
 * No. And rightly so. 172.254.255.250 (talk) 18:52, 31 July 2019 (UTC)
 * at can be used for this purpose. --Izno (talk) 22:39, 31 July 2019 (UTC)

Partial links in title
According to the docs, title accepts wikitext is title-link is not set. However, the code is mistakenly rendered with the title Gerhardt. This is because is_wikilink, which is called from kern_quotes ignores text before and after the wikilink. I was wondering if this is intentional or a bug?--Strainu (talk) 18:12, 30 July 2019 (UTC)
 * I have moved this dicussion because this is a more public forum and where we handle bug reports.
 * —Trappist the monk (talk) 18:58, 30 July 2019 (UTC)
 * I would assume the behavior described in the OP is intentional. Linking of the title/work in citations is supposed to provide additional info about the source itself, it is not meant to convey any other info, including info about the work's subject. 65.88.88.69 (talk) 19:07, 30 July 2019 (UTC)


 * Likely a bug though I would also argue that wikilinking as you show in your example really isn't appropriate because a link to an article about Gerhardt isn't going to do much to help readers locate a copy of an article (by Marc Tiffeneau, right? comma missing in author?).  If Gerhardt is important enough to be wikilinked, then a wikilink from the article's prose is a much better place for that link.
 * Further, while the title is truncated by this bug, it is not italicized as you show in your post:
 * Still, cs1|2 shouldn't corrupt source titles by truncating to the piped link; un-piping the wikilink appears to work:
 * so I'll look into it.
 * —Trappist the monk (talk) 19:17, 30 July 2019 (UTC)
 * Before I get too involved with 'fixing' this, should I? The truncation issue is I think fixed:
 * As it works now,  correctly handles kerning when title holds a value that is one of these forms:
 * 'plain text' →
 * 'en.wiki article title' →
 * 'label' →
 * 'label' →
 * 'plain text en.wiki article title ' →
 * ' en.wiki article title plain text' →
 * 'plain text en.wiki article title plain text' →
 * 'plain text label plain text' →
 * it doesn't properly handle these two mixed plain text / wikilink when the wikilink(s) begin or end the title:
 * plain text 'label' →
 * 'label' plain text →
 * The question is, should it handle 5–8 and the two that it currently doesn't, or should mixed plain text / wikilink titles throw an error? Are there cases where partially wikilinked titles help readers locate the source?  Certainly an article about a book when there isn't a courtesy url might be helpful but that should be the whole title wikilinked.  Mixed plain text and wikilinks in chapter (and aliases)?  in title for the periodical templates ?
 * —Trappist the monk (talk) 12:57, 1 August 2019 (UTC)
 * I wouldn't bother. Titles are wikilinked because they refer to sources. We verify sources according to the complete, correct titles. I cannot see where a title-fragment would even be allowed in a citation as a verifiable source. Linking a fragment just because there's a Wikipedia article for it does not offer anything for citation purposes imo, and could lead to charges of OR when used to divine a title. But even if it did offer an advantage, there's way too much complexity (obvious by your questioning above) in exchange for dubious results. 65.88.88.217 (talk) 17:42, 1 August 2019 (UTC)
 * I have seen titles along the lines of Star Wars: A New Hope in citations. One of the current naming conventions (WP:SUBTITLES) indicates that some works have their titles trimmed on Wikipedia. I also do not know if that should be a valid use, or if we should error when the link does not include the entire part of the work's name. --Izno (talk) 03:38, 2 August 2019 (UTC)
 * it doesn't properly handle these two mixed plain text / wikilink when the wikilink(s) begin or end the title:
 * plain text 'label' →
 * 'label' plain text →
 * The question is, should it handle 5–8 and the two that it currently doesn't, or should mixed plain text / wikilink titles throw an error? Are there cases where partially wikilinked titles help readers locate the source?  Certainly an article about a book when there isn't a courtesy url might be helpful but that should be the whole title wikilinked.  Mixed plain text and wikilinks in chapter (and aliases)?  in title for the periodical templates ?
 * —Trappist the monk (talk) 12:57, 1 August 2019 (UTC)
 * I wouldn't bother. Titles are wikilinked because they refer to sources. We verify sources according to the complete, correct titles. I cannot see where a title-fragment would even be allowed in a citation as a verifiable source. Linking a fragment just because there's a Wikipedia article for it does not offer anything for citation purposes imo, and could lead to charges of OR when used to divine a title. But even if it did offer an advantage, there's way too much complexity (obvious by your questioning above) in exchange for dubious results. 65.88.88.217 (talk) 17:42, 1 August 2019 (UTC)
 * I have seen titles along the lines of Star Wars: A New Hope in citations. One of the current naming conventions (WP:SUBTITLES) indicates that some works have their titles trimmed on Wikipedia. I also do not know if that should be a valid use, or if we should error when the link does not include the entire part of the work's name. --Izno (talk) 03:38, 2 August 2019 (UTC)

"developer=" ?
I have just recently encountered cites with the "unsupported parameter" developer. Is there some discussion about adding this parameter? (I hope not) -- User-duck (talk) 06:34, 4 August 2019 (UTC)

Seems to be related to: Replaced Infobox online music service with Infobox online service. -- User-duck (talk) 07:27, 4 August 2019 (UTC)
 * Way back, I think before citation templates were standardized, I think there was . With a "developer" parameter. 98.0.145.210 (talk) 12:26, 4 August 2019 (UTC)
 * cite video game has a developer, but I do not think any other templates employ it. --Izno (talk) 14:56, 4 August 2019 (UTC)

Tracking bad et al
I just found [//en.wikipedia.org/w/index.php?title=Special:Search&limit=250&offset=0&ns0=1&search=insource%3A%2F+\.\.\.+%5B\%5B+%5D*et%5B+%5D*\|%2F&advancedSearch-current={} about 170] incorrect parsing of the et al in an author list (possibly by a script). if I have time, I may try to fix some of these, but it would be good to have some tracking to assist (if possible). thank you. Frietjes (talk) 18:48, 9 August 2019 (UTC)
 * Some of them are identifiable by the other author parameter as well, e.g. al.. That one seems harder to track, but the your pattern might be reasonably possible. (Though, otoh, maybe just in first; author is used to hold pseudonyms, which sometimes include brackets.) --Izno (talk) 05:56, 10 August 2019 (UTC)

DOI check, 10.5555 = test doi
All 10.5555 DOIs are test DOIs and will never resolve. There should be an error thrown for those. &#32; Headbomb {t · c · p · b} 18:50, 9 August 2019 (UTC)
 * In the sandbox:
 * —Trappist the monk (talk) 11:07, 10 August 2019 (UTC)
 * —Trappist the monk (talk) 11:07, 10 August 2019 (UTC)
 * —Trappist the monk (talk) 11:07, 10 August 2019 (UTC)

doi-broken-date without doi
gives

The second case should be in a tracking category so that doi-broken-date can be removed. &#32; Headbomb {t · c · p · b} 16:14, 9 August 2019 (UTC)
 * Agreed, this should be an error. --Izno (talk) 05:56, 10 August 2019 (UTC)
 * In the sandbox:
 * Categorizes to.
 * —Trappist the monk (talk) 11:06, 10 August 2019 (UTC)
 * It's putting a lot of weird things in Category:CS1 errors: DOI. A namespace check is probably needed. &#32; Headbomb {t · c · p · b} 12:10, 10 August 2019 (UTC)
 * fixed.
 * —Trappist the monk (talk) 12:58, 10 August 2019 (UTC)
 * It's putting a lot of weird things in Category:CS1 errors: DOI. A namespace check is probably needed. &#32; Headbomb {t · c · p · b} 12:10, 10 August 2019 (UTC)
 * fixed.
 * —Trappist the monk (talk) 12:58, 10 August 2019 (UTC)

Also throws a weird error As do a lot of things in .&#32; Headbomb {t · c · p · b} 12:11, 10 August 2019 (UTC)
 * I don't see anything strange in.
 * —Trappist the monk (talk) 12:58, 10 August 2019 (UTC)
 * The fix above got rid of those errors as well. &#32; Headbomb {t · c · p · b} 13:17, 10 August 2019 (UTC)

Chapter + Contributor = Error
I find myself here from 's talk; Using both chapter and contributor* throws an error. Am I doing something wrong, or more importantly, how do I fix this? 06:11, 12 August 2019 (UTC)
 * Are you citing something Mitchell said in the foreword, of something Warner said in the body of the book? If you're citing the book, there's no need to mention the forword in a citation.  It's also unnecessary to specify the chapter if the whole book is by the same author.  If you are citing both, you need separate citations.  Kanguole 07:26, 12 August 2019 (UTC)
 * Contributors and chapters are not the same thing and should not be an alias of each other. I'll figure something out to workaround this error, but it should be fixed.  18:09, 12 August 2019 (UTC)
 * Umm, contributor and chapter are not aliases of each other. However, contribution and chapter are aliases.
 * The value in url from your examples takes readers to, for lack of a better term, a dab page at books.google.com. Neither of the two offerings there are part of Mitchell's foreword.  Since it would appear that you are not citing Mitchell's foreword, she and the foreword should not be part of the citation.  The error message is correct.
 * —Trappist the monk (talk) 18:34, 12 August 2019 (UTC)
 * So the contribution should only be mentioned if it's being directly cited? Fair enough; I see that now; thanks. Since I'm sure it would not be trivial to output a special error that better explains its self in this case, perhaps a clear note in either the template docs or at the CS1 error help would be a good idea. Being informed that there are repeated params that are not repeated, without explanation, is just plain annoying and lead to this need for clarification.  18:46, 12 August 2019 (UTC)
 * So the contribution should only be mentioned if it's being directly cited? Fair enough; I see that now; thanks. Since I'm sure it would not be trivial to output a special error that better explains its self in this case, perhaps a clear note in either the template docs or at the CS1 error help would be a good idea. Being informed that there are repeated params that are not repeated, without explanation, is just plain annoying and lead to this need for clarification.  18:46, 12 August 2019 (UTC)

Problem with URL containing non-Latin characters
I have a URL that is throwing an error, "https://нэб.рф/catalog/000200_000018_RU_NLR_DIGIT_11048/viewer". The URL works (takes one to the desired location on the web), but it throws an error, and coming here I see "The URL field is checked to ensure that it contains only Latin characters..."

So after looking it up, I substited the "encoded" version, which is ""%D0%BD%D1%8D%D0%B1.%D1%80%D1%84", in place of the Cyrillic characters "нэб.рф", but it still throws an error, altho this version also works.

So my questions are: 1) Why throw an error when it works? 2) What's the solution?

(As with all templates, I wish there was a "don't check, I know what I'm doing" field, but there isn't AFAIK.) Herostratus (talk) 14:59, 12 August 2019 (UTC)


 * This is an internationalized domain name (IDN), with an added top-level domain implementation because of the рф portion. These use IDNA encoding ie. punycode, not percent-encoding. Thus нэб.рф converts to the plain-text but odd looking xn--90ax2c.xn--p1ai (converter tool) which actually works: http://xn--90ax2c.xn--p1ai .. the question is should the punnycode conversion be the responsibility of the editor adding the domain; or can CS1|2 accept IDNs; or should a bot search out and convert them to punycode when in CS1|2.  --  Green  C  15:13, 12 August 2019 (UTC)
 * When I first added support for punycode detection to cs1|2 I though about encoding. At the time, unicode urls were quite rare so I did not pursue that.  Given the number of complaints about them (few), it would seem that they are still rare or that editors don't mind doing the encoding.  If we do decide to accept and encode unicode urls, it would seem that cs1|2 is where it should be done.  That way the unicode url remains with the source (though it can, of course, be decoded).
 * —Trappist the monk (talk) 18:44, 12 August 2019 (UTC)
 * Oh, OK, this works, thank you. Yeah I mean trying to figure this out, failing to find anything right off on the internet, trying the unicode, figuring out where to ask for help, and all that, tool me an hour at least. So a bot doing that would save x manhours over time. Or something in the documentation.


 * And I mean, as I said, the Cyrillic version works. Something somewhere is able to interpret it and take the reader to the desired place. It seems kind of like "Well, the check engine light comes on whem the framitz is out of alignment, but the car works just as well anyway". Well then why did I make this trip to the dealer. I suppose maybe it's hygiene thing, where you don't want to ask whatever interprets the URL to do more work that you need to, maybe in case whoever controls that decides they don't want to support that function anymore, or something. But OK. It's not a big deal, and I myself am now gruntled. Thans again. Herostratus (talk) 19:08, 12 August 2019 (UTC)
 * Agree CS1|2 should be able to parse an IDN without warning/error:
 * has about 6,000 pages (most of which are likely not caused by IDN) so it's not a big problem but would be a nice feature to support. -- Green  C  23:34, 12 August 2019 (UTC)
 * has about 6,000 pages (most of which are likely not caused by IDN) so it's not a big problem but would be a nice feature to support. -- Green  C  23:34, 12 August 2019 (UTC)

pages parameter
Pages are displayed ":57–59", but the explanation says they are displayed "pp. 57–59". Am I doing something wrong? Template:Cite magazine displays pages "pp. 57–59". Vzeebjtf (talk) 22:20, 12 August 2019 (UTC)
 * You don't give an example of what you are doing so I have to assume that you're using in which case 57–59 displaying as ":57–59" is correct.  Similarly, using  with the same 57–59 displaying as "pp. 57–59" is also correct.  Give an example of what you are doing.
 * —Trappist the monk (talk) 22:28, 12 August 2019 (UTC)
 * I'm sorry -- I was sloppy in asking my question. See Template:Cite journal/doc. It says:"page: The number of a single page in the source that supports the content. Use either"—which does not seem to be accurate. Perhaps it should be corrected? Vzeebjtf (talk) 23:42, 12 August 2019 (UTC)
 * I was confused by this question until I realized that the cite journal documentation for the page and pages parameter merely calls csdoc (the documentation for Citation Style 1 more generally) and that this general-purpose documentation does not accurately describe the journal-specific treatment of pages. I'm not sure of the correct fix: parameterize csdoc to allow it to know that it should use the journal-specific page description, or expand it for that section in cite journal and then correct the expansion? Probably parameterizing the template would be more flexible for future changes. On the other hand I wouldn't be strongly opposed to changing CS1 so that journals are more in line with everything else in this respect. —David Eppstein (talk) 00:55, 13 August 2019 (UTC)
 * Cite journal has many deviating aspects and AFAICS it's due to no-one trying to change it, because it comes up often enough how its various deviations are deviations.... :) --Izno (talk) 01:08, 13 August 2019 (UTC)

Protected edit request on 19 August 2019
Under, in the examples for use of the  parameter, please add box and table (e.g. "Box 8.1" and "Table 8.1" on page 195 of ACSM's Guidelines for Exercise Testing and Prescription). Thank you. Sangdeboeuf (talk) 01:05, 19 August 2019 (UTC)
 * cs1|2 documentation is not fully protected; you can make the changes yourself.
 * —Trappist the monk (talk) 01:18, 19 August 2019 (UTC)
 * —Trappist the monk (talk) 01:18, 19 August 2019 (UTC)

Error - Bibcode Journal (JJJJJ = E3SWC) contains a number
I searched the archives but haven't seen this mentioned. The bibcode check assumes that the JJJJJ will only have letter, ampersand, or dot, but 'E3S Web of Conferences' = E3SWC breaks the standard and throws an invalid bibcode error (see Food_safety_incidents_in_China); Can the templates be updates to validate this? Quuux (talk) 01:03, 23 August 2019 (UTC)

fixed in the sandbox:

—Trappist the monk (talk) 10:48, 23 August 2019 (UTC)

website or work?
I wanted to ask which is preferred, I've noticed =work being changed to =website by some users. Govvy (talk) 16:18, 24 August 2019 (UTC)
 * In Template:Cite web, work is an alias of website; there is no reason to change it. I don’t think one is preferred over the other. Umimmak (talk) 17:06, 24 August 2019 (UTC)