Help talk:Citation Style 1

Citation templates

... in conception

... and in reality

[Template:Cite book] Contributions not working?

I cannot seem to find a way to add authors of prefaces or forewords. According to the documentation at Template:Cite book, this is done by the parameters contribution, contributor-last and contributor-first. However, those parameters do not seem to work and are not present in the Visual Editor's interface for Template:Cite book.
Is there a bug is did I fail to understand how it is supposed to work? Veverve (talk) 17:40, 17 December 2021 (UTC)

It is always good to provide a non-working example.

{{cite book |title=Title |contribution=Foreword |last=Smith |first=CD |contributor-last=Jones |contributor-first=AB}}

Jones, AB. Foreword. Title. By Smith, CD.

The above appears to work. What are you attempting to do that does not work?

—Trappist the monk (talk) 18:01, 17 December 2021 (UTC)

Are you citing the preface or forward? That's what |contribution=, |contributor= / |contributor-last=, etc., is for. For instance, if you are citing John Smith's statement that "working with Jane Doe was a labor of love",, you'd use the following citation:

• {{Cite book |contributor-last=Smith |contributor-first=John |contribution=Forward |last=Doe |first=Jane |date=2021 |title=Fake Book I Made Up }}
• Smith, John (2021). "Forward". Fake Book I Made Up. By Doe, Jane.

You can see that the citation is not quoting Jane Doe in her book, it's quoting the person who wrote the contribution (similar to an individual author's article in a collaborative encyclopedia, {{Cite encyclopedia}} cite the article entry & author, not the editor of the encyclopedia)

 — sbb (talk) 18:08, 17 December 2021 (UTC)

References

@Sbb: I am trying to add: Ward, Gary L.; Persson, Bertil; Bain, Alan, eds. (1990). "Brown, Daniel Quilter". Independent Bishops: An International Directory. Apogee Books. p. 63. ISBN 978-1-55888-307-9. {{cite book}}: |contributor= requires |author= (help); |contributor= requires |contribution= (help); More than one of |contribution= and |chapter= specified (help). Veverve (talk) 18:22, 17 December 2021 (UTC)

(edit conflict)

In that you wrote:

|chapter=Brown, Daniel Quilter

|contribution=Preface

When cs1|2 see two parameters that are aliases of each other, it chooses one of them and emits the More than one of |contribution= and |chapter= specified error message. Because cs1|2 chose |chapter=, it ignores |contribution=. Because you wrote:

|contributor-last=J. Gordon |contributor-first=Melton

and because cs1|2 is ignoring |contribution= (which is required for any of the |contributor= parameters), cs1|2 emits the |contributor= requires |contribution= error message. Because there is no |author= (or |last=) parameter (which is required when using any of the |contributor= parameters, cs1|2 emits the |contributor= requires |author= error message.

You cannot cite two (or more) sections of a book in a single {{cite book}} template. If you want to cite both the chapter "Brown, Daniel Quilter" and the "Preface", you must do so separately. And, since this source appears to be more of an encyclopedia than a book with chapters (to me, "Brown, Daniel Quilter" 'feels' more like an encyclopedia entry) then, for that entry, perhaps this:

{{Cite encyclopedia |url=https://books.google.com/books?id=EpXjAAAAMAAJ&q= |encyclopedia=Independent Bishops: An International Directory |date=1990 |publisher=Apogee Books |isbn=978-1-55888-307-9 |editor-last=Ward |editor-first=Gary L. |page=63 |language=en |entry=Brown, Daniel Quilter |editor-last2=Persson |editor-first2=Bertil |editor-last3=Bain |editor-first3=Alan}}

Ward, Gary L.; Persson, Bertil; Bain, Alan, eds. (1990). "Brown, Daniel Quilter". Independent Bishops: An International Directory. Apogee Books. p. 63. ISBN 978-1-55888-307-9.

and for the preface, perhaps this:

{{Cite encyclopedia |url=https://books.google.com/books?id=EpXjAAAAMAAJ&q= |encyclopedia=Independent Bishops: An International Directory |date=1990 |publisher=Apogee Books |isbn=978-1-55888-307-9 |editor-last=Ward |editor-first=Gary L. |page=63 |language=en |section=Preface |editor-last2=Persson |editor-first2=Bertil |editor-last3=Bain |editor-first3=Alan}}

J. Gordon, Melton (1990). "Preface". In Ward, Gary L.; Persson, Bertil; Bain, Alan (eds.). Independent Bishops: An International Directory. Apogee Books. p. 63. ISBN 978-1-55888-307-9.

—Trappist the monk (talk) 18:55, 17 December 2021 (UTC)

So again, what are you trying to cite/quote? Are statements made by J. Gordon Melton in the Preface being cited, or is the chapter/entry "Brown, Daniel Quilter" being cited? If the answer is both, then you either need 2 different {{Cite xxx}} templates, or if your article uses Harvard-style citations, you can use {{harvc}} to compactly cite and markup multiple entries in the same collection. If the Preface isn't being cited at all, just leave it out of the citation.

Note: For what you're doing, I'd use {{Cite encyclopedia}} over {{Cite book}}. c.f.:

With |contribution=

Ward, Gary L.; Persson, Bertil; Bain, Alan, eds. (1990). "Preface". Brown, Daniel Quilter. Independent Bishops: An International Directory. Apogee Books. p. 63. ISBN 978-1-55888-307-9. {{cite encyclopedia}}: |contributor= ignored (help)

Without |contribution=

Ward, Gary L.; Persson, Bertil; Bain, Alan, eds. (1990). "Brown, Daniel Quilter". Independent Bishops: An International Directory. Apogee Books. p. 63. ISBN 978-1-55888-307-9.

Notice with the contribution, the "Preface" is quoted (as a dictionary/encyclopedia article should be), whereas the article entry is italicized? I think the contribution is confusing the {{Cite xxx}} code.  — sbb (talk) 18:41, 17 December 2021 (UTC)

(edit conflict)

Your first example attempts to cite both "Preface" and "Brown, Daniel Quilter" which cs1|2 really isn't designed to do. One source per template.

—Trappist the monk (talk) 19:03, 17 December 2021 (UTC)

@Sbb: I am trying to indicate that this book has a preface by J. Gordon Melton, and that the entry "Brow, Daniel Quilter" is being cited.

{{Cite encyclopedia}} would not be very useful in this case, as there is no indication as to which contributor wrote which entry inside the book. Veverve (talk) 18:58, 17 December 2021 (UTC)

One source per cs1|2 template. If you are including "Preface" in the citation merely because the work has a preface, don't do that. If you are citing something from "Preface", use a separate cs1|2 template and, I agree with Editor sbb that {{cite encyclopedia}} appears to be the best choice here.

—Trappist the monk (talk) 19:03, 17 December 2021 (UTC)

@Trappist the monk: If you are including "Preface" in the citation merely because the work has a preface, don't do that. I think it is quite relevant to indicate if there is a preface to a work, which is why I tried to indicate it. It is especially important in this case, as J. Gordon Melton is a well respected encyclopedia editor in the religious field. Veverve (talk) 19:08, 17 December 2021 (UTC)

If you are citing the preface, cite the preface. The purpose of a citation is to help readers locate the source that supports the text in an en.wiki article. If the text at Daniel Q. Brown is supported by the source named in the current citation (the "Brown, Daniel Quilter" entry) then that is all that is needed. If the preface is not being used to support text at Daniel Q. Brown, tt really does not matter that J. Gordon Melton is a well respected encyclopedia editor. When/if text at Daniel Q. Brown is supported by Melton's preface, then cite Melton separately.

You might write |others=Preface by J. Gordon Melton. I do not recommend it because that is just so much extraneous text that doesn't aid a reader in locating a copy of the source.

—Trappist the monk (talk) 19:42, 17 December 2021 (UTC)

Mutually agreeing with Trappist the monk here. I get it, and sympathize with your goal from a "data completionist" standpoint. The information is there, why not include it, right? But please understand, that's not what the {{Cite xxx}} or {{Citation}} templates are really meant for. They are meant to support citations supporting references in en.wiki articles. The |contribution= and related parameters exist in support of citing material in the text. Think of it this way: if articles needed to cite a Preface or Forward of a work, how would they do so, given the current crop of {{Cite xxx}} templates, if the templates worked the way you're trying to make them work? The answer is, they couldn't. That's what |citation= |contribution= et. al provide: a way to cite material that's not part of the main body of the work.

BTW, it doesn't matter that there's no specific author that can be identified for the "Brown, Daniel Quilter" article. You should still be using {{Cite encyclopedia}}. The work you're citing is Independent Bishops: An International Directory. From the introductory sentence of the template's doc page:

This ... template is used to create citations for articles or chapters in edited collections such as encyclopedias and dictionaries, but more generally any book or book series containing individual sections or chapters written by various authors, and put together by one or more editors.

If you want to note additional facts about the citation, just add it after the template, and use |mode=cs2 as an argument to suppress the final stop/period that the {{Cite encyclopedia}} template normally provides:

• {{Cite encyclopedia |mode=cs2 |title=Brown, Daniel Quilter |encyclopedia=Independent Bishops: An International Directory |date=1990 |publisher=Apogee Books |isbn=978-1-55888-307-9 |editor-last=Ward |editor-first=Gary L. |pages=63 |editor-last2=Persson |editor-first2=Bertil |editor-last3=Bain |editor-first3=Alan |url=https://books.google.com/books?id=EpXjAAAAMAAJ&q= }}, with Preface by [[J. Gordon Melton]].
• Ward, Gary L.; Persson, Bertil; Bain, Alan, eds. (1990), "Brown, Daniel Quilter", Independent Bishops: An International Directory, Apogee Books, p. 63, ISBN 978-1-55888-307-9, with Preface by J. Gordon Melton.

edited: added |ref=none to the {{Cite encyclopedia}} template to suppress the 'Harv warning' in this example. If you are using CITEREFs / reference templates ({{sfn}}, {{harvnb}}, etc.), then you don't need the |ref=none parameter to the Cite template.  — sbb (talk) 02:57, 18 December 2021 (UTC)

Umm, quibbles:

• no such parameter as |citation=
• |mode=cs2 also switches the element separator from dot to comma and changes some capitalization so templates with that parameter/value pair are stylistically different from adjacent cs1 templates. This can cause knickers to twist. Better perhaps is to use |postscript=none or |postscript=, (or other single punctuation character).

—Trappist the monk (talk) 12:16, 18 December 2021 (UTC)

Dang it. Thanks for catching my |citation= mistake (I meant to say |contribution=, but I had {{Citation}} on my mind from the previous sentence). Edited with strike and ins.

And I should have known I was being a bit too cavalier with |mode=cs2, but I forgot it had other effects besides just final punctuation. Definitely agree, I shouldn't recommend CS1/CS2 mode switching; people go to a lot of effort to get articles self-consistent. Thanks.  — sbb (talk) 18:45, 18 December 2021 (UTC)

Publication date for journal references

{{cite journal}} accepts both year and date, but the former is discouraged. Since date can accept day, month, and year; this creates inconsistency in the references. I understand the full date is necessary for some references such as news articles, but it is really superfluous for scholarly journals. I suggest encouraging date=YYYY for journal references for the following reasons: (i) In scholarly referencing, the full date of publication is (almost) never used for journal articles. (ii) Full date of publication is technically useful for scholarly articles, as the dates of submission, acceptance, online publication, print publication are months apart in the best-case scenario. (iii) The guide says it is useful if an author publishes several articles in the same year, but Wikipedia references are numbered not sorted by authors' names in Chicago style. Even in the latter, YYYYa,b,c is used since there is no guarantee that one author has not several articles in the same month.

My suggestion is not to change anything but to encourage in the referencing guide using the year of publication.589q (talk) 02:25, 23 December 2021 (UTC)

The date used in journal citations reflects the date of publication of the specific issue. This has to be so because this is one of the fields indexed by periodical reference/biblio database providers. Therefore it is an important aid in finding the source of the citation and verifying the wikitext. Wikipedia by its very nature demands this. It should not be compared to other projects that have different audiences, specifications, or demands. 71.247.146.98 (talk) 12:50, 23 December 2021 (UTC)

(i) Nobody uses the issue to find an article. Every journal article is linked with DOI. (ii) 80% of journal articles in Wikipedia references are cited by year only. My suggestion is for consistency. (iii) The reset usually have a YYYY-MM format, which is not enough for finding an issue because most journals have more than 12 issues per year. (iv) Most journals do not even mention the publication date in the list of issues (only issue number).589q (talk) 13:45, 23 December 2021 (UTC)

It is not correct to say that every journal article is linked with a DOI. Not all of them are, for many different reasons. Secondly, you assume that the average Wikipedia reader would know what a DOI is, as it is generally considered scholarly/expert info. As a matter of course, registrants such as CrossRef assign part of the suffix by journal date and/or issue. Also, it doesn't matter how journal references are cited in Wikipedia; we are only discussing how they should be correctly cited. A reader wanting to verify a citation can see several non-cryptic pieces of information before the identifiers. This order of placement is not random. As noted before, reference databases index author/date/issue/article title/identifier per journal. That is the easiest way for the average reader to find them. Not coincidentally, DOI registrants may use the very same databases (among other information) to build DOI suffixes. The resulting DOI number will then be used to update the relevant databases. 65.88.88.68 (talk) 15:30, 23 December 2021 (UTC)

You got me wrong on the DOI matter. I didn't say people use DOI to check the citations. I said (almost) every journal article referenced in Wikipedia has a hyperlink via doi.org which sends the reader to the article. For checking the reference, people simply click on the available hyperlink. Nobody finds the article via the table of contents when there is a direct link, whether you are familiar with DOI or not. All scholarly articles have DOI. My suggestion is all about the readability for both professional and average readers. I just suggest consistency for "Doe, J. (2021)", "Doe, J. (December 2021)", and "Doe, J. (23 December 2021)" referencing. Roughly speaking, 80% of journal articles in Wikipedia references use the first format. I just suggest having a recommendation for the preferred format. When I add a reference, I do not know which format is recommended. Thus, I pick one randomly. These random choices add to the unnecessary inconsistency.589q (talk) 17:37, 23 December 2021 (UTC)

Not trying to prove anybody wrong on anything, we are discussing a proposition. I know that |date= documentation suggests several different ways to input the date. Presumably, the source can be found whether the date is input as year only or as a more complete date. However it is best to input the date the way you see it in the publication itself, again because this is how the source will be found faster when one searches for it by date. The consistency proposition satisfies aesthetics but may hamper/delay discovery. In an inherently unreliable project like Wikipedia, ease of source discovery and therefore of verification is important. Consistency regarding dates is recommended where it will not affect discovery. And may I repeat that the idea that the average reader will see a citation and immediately click on an otherwise cryptic identifier is not a useful assumption, and imo highly unlikely. 65.88.88.68 (talk) 17:06, 23 December 2021 (UTC)

I am probably on a different page, as I have no clue what you are referring to. "this is how the source will be found faster when one searches for it by date". Could you please let me know how you search for a journal article by the date of publication? And when there is a direct hyperlink to the article, why should one search for the article?589q (talk) 18:28, 23 December 2021 (UTC)

Not all of the references that are cited in Wikipedia are online, and some readers may have to resort to searching through physical journals at libraries (particularly for older works), in which case details of how the journal refers to the location etc of the article (i.e. how the journal presents the date of the issue and things like volume and issue number, and even page numbers) will be very helpful in finding the article in question. Your proposal merely strengthens FUTON bias.Nigel Ish (talk) 20:42, 23 December 2021 (UTC)

Even when the full text is available online, you may still have to search on the date, drilling down to the year, month, volume and issue. This occurs when the journal has been digitised by being scanned, but the text is not searchable. Hawkeye7 (discuss) 21:16, 23 December 2021 (UTC)

(i) I have not come across a single journal article without a PDF version. If there is, it is almost impossible to find them in any library. (ii) If someone has access to a library with a physical archive of the scholarly journals, the library should have an online subscription too. (iii) I come from the generation of hardcopy and spent nights in libraries for finding references. I have never used the publication date for finding an article because it is technically impractical. Journal issues are bonded together and only the range of issues is printed on the cover. (iv) If I am wrong, and the date of publication is useful in finding an article, then, there should be a recommendation for providing the full date of publications for the 80% of the Wikipedia references, which have provided year only.

My suggestion was simple. I asked when I am adding a reference, which of the three formats (bolded above) is recommended? If your answer is: whichever you like, it is not appropriate for an encyclopedia. If your answer is: whichever is available, I have the full date of thousands of (almost all) references with year only. Should we add the missing information? If your answer is: leave them the way they are, it is not appropriate for an encyclopedia because the choice was based on personal preferences to include the full date or year only. 589q (talk) 00:27, 24 December 2021 (UTC)

589q, please provide data for your claims about date formats and the presence of DOIs for journal citations in WP articles. Your claims are wildly inconsistent with my experience in editing tens of thousands of citations in articles. Thanks. – Jonesey95 (talk) 00:59, 24 December 2021 (UTC)

Jonesey95, I didn't get what part was different from your experience. Do you mean the DOI is not available for most journal articles cited in Wikipedia? Or the date formats are not mostly years? For example, see Lithium as a general article with numerous similar ones.589q (talk) 01:29, 24 December 2021 (UTC)

[edit conflict] The answer really is: whichever you like. I tend to use either year=2021 or date=December 2021; I don't think the full dates are useful, but the months sometimes are, for the purpose suggested above: tracking down articles in journals from the journal name and publication date in cases where the title isn't working well. Re the discussion above about journal publications with no online pdf: they're unusual but not unheard of. For instance I don't know where to find the following in its original form (I do know of a revised version in a 2019 book that is available online): Datta, Bibhutibhusan; Singh, Awadhesh Narayan (1992). "Use of permutations and combinations in India". Indian Journal of History of Science. 27 (3): 231–249. MR 1189487. Re your attitude that there can only be one format: See WP:CITEVAR. —David Eppstein (talk) 01:04, 24 December 2021 (UTC)

David Eppstein, it is very unlikely to find that article in physical form in libraries too. Does it help to find it easier? Datta, Bibhutibhusan; Singh, Awadhesh Narayan (24 December 1992). "Use of permutations and combinations in India". Indian Journal of History of Science. 27 (3): 231–249. MR 1189487..589q (talk) 01:36, 24 December 2021 (UTC)

The issues I can find online of that journal [1] are labeled by month and year, but not by the day within the month. So the month and year are useful information to me; the day is not. Also, the issue with that paper appears to be available in a physical copy at the Stanford University library, very near where I happen to be now and where (as an alumnus) I would probably have access to it if it weren't a holiday, and I definitely would have access to photocopies of either that or the UC Berkeley copy through interlibrary loan from my home library at UC Irvine. —David Eppstein (talk) 01:44, 24 December 2021 (UTC)

Normally, when looking up for an article in the physical archive, volume and page number is sufficient. Extra information is the issue number. The month of publication does not provide any information beyond the issue number. The issue number is usually mentioned in both the academic bibliography and Wikipedia reference (including your example). If we want to add more information, why not using one that is available, common, and more accurate (issue number), and just add extra data, which might be useful or not. By the way, if you travel to another institution (whatever close) to get an article, you wouldn't care to have the month of publication in hand to find it 10 seconds faster.589q (talk) 02:00, 24 December 2021 (UTC)

Point of note, "YYYY-MM" is not an acceptable date format. This is because it is ambiguous if a date such as "2006-07" refers to July 2006, or the year range "2006–2007".  — sbb (talk) 16:17, 24 December 2021 (UTC)

The statement "if date is useful in finding a journal article" is incorrect. Journal dates are indexed by all metadata providers so that search queries can find a particular issue by date. There is no "if" here, this is one of the items that sources are classified by. So date has semantic significance in the context, and is not just a matter of aesthetics. It is also not an obtuse item for the average reader, unlike any identifier. That said, nobody forces anybody to use the complete issue date (even though they should). But it would be semantically diminishing to suggest a date abbreviation as the "preferred" date. If anything, the complete date should be suggested as the preferred option. 69.203.140.37 (talk) 22:23, 24 December 2021 (UTC)

Clarification: above, it was not meant that editors should be obligated to use the complete publication date as it appears in the journal. It is just my opinion that they should, for the sake of faster/easier discovery. 172.254.162.90 (talk) 23:43, 24 December 2021 (UTC)

Almost everyone is in favor of the month of publication. I still believe an article is looked up by its year, volume, and page number if DOI is not available. I just wonder if the month of publication is such useful, why is there no petition to add the missing date of publication in the references. I come to my second reasoning that the date of publication is misleading.

Let me clarify this through an example. How do you cite this article https://pubs.acs.org/doi/10.1021/jacs.1c08484? At the top of the page, it is written: Publication Date:November 28, 2021, but this article belongs to volume 143, issue 49 (December 15, 2021). Nowhere on the page, you can find December 15, 2021 unless you go to the issue page.

You may wonder what if the publication year is different from the issue. Here is an example: https://pubs.acs.org/doi/10.1021/jacs.0c10943 Publication Date:December 15, 2020 belongs to volume 143, issue 1 (January 13, 2021). At the top of the page, it is clearly stated Cite this: J. Am. Chem. Soc. 2021, 143, 1, 5–16

The problem is: for most publishers, the publication date is the date of (online) publication of the article, but the publication year is of the release of the whole issue. The latter must be used for citation. Usually, the latter is included in the former, but not necessarily (the second example I gave). 589q (talk) 03:26, 26 December 2021 (UTC)

My mistake: I should have used "issue date" where I used "publication date". Most (not all) journals have a set issue date schedule, but often the publication schedule diverges for whatever reason. For this reason, most biblio providers classify by "issue date" if it exists, rather than "publication date", which comes into play if it is the only date. If you look at the way providers structure the metadata, "date" almost always refers to issue date (usually a known in-advance property), with publication date commonly relegated to a "Notes"-like field. In CS1, a citation can provide both dates, although there is an opinion to do away with publication date, and use that info only when issue date is absent (confining this argument to journals for now).

In you first example, suppose one remembers part of an author's/editor's last name. It is also known that the work was published in late 2021. With that info, you will get a list of results. I suggest that in the great majority of queries the entry with the author and issue date will appear before any other date in a browser with no previously cached results, simply because that is how the info is indexed, and search engines use this same info. If you know the journal name, it is easier, because the issue schedule is set, and you can zero-in on the "December 2021" issue. And as you can see in the second example this correct procedure is given (cite by issue date), as this leads to the most efficient doscovery. 68.173.76.118 (talk) 14:25, 26 December 2021 (UTC)

Reiterating that this discussion is about serial publications. Other considerations apply in e.g. books, including book series. 68.173.76.118 (talk) 14:42, 26 December 2021 (UTC)

We have two independent parameters: |date=, for the date printed on the cover; and |publication-date= for the date that it became available to subscribers and other potential purchasers. These two need not be the same, see Template:Cite journal#csdoc_date and Template:Cite journal#csdoc_publication-date. For example:

• Feifan Wang; Yongping Fu; Ziffer, Mark E.; Yanan Dai; Maehrlein, Sebastian F.; X.-Y. Zhu (January 13, 2021). "Solvated Electrons in Solids—Ferroelectric Large Polarons in Lead Halide Perovskites". Journal of the American Chemical Society. 143 (1) (published December 15, 2020): 5–16. doi:10.1021/jacs.0c10943.

Is that way of showing both dates satisfactory? --Redrose64 🌹 (talk) 00:07, 27 December 2021 (UTC)

It seems clear to me that the positer of this query likely doesn't write historical articles nor possibly on foreign topics. As someone who does, I can definitely state that it is less often than frequent that there is a DOI reference to the majority of articles I use as reference materials. Unequivocally knowing the date as well as volume and issue are incredibly important. Many journals which have existed for decades have altered their range numbers, thus you must tie both the date of the issue to the vol/no. I cannot even estimate how many times I have used the "ask a librarian" feature of world cat to retrieve and have e-mailed to me an article that exists in a publication that has not been digitized. In every single instance, I was asked to provide the issue date as well as the volume and issue number. While DOI may be common in the US/UK, it is less so in the rest of the world mainly because it requires that sources be digitized, which is much less frequent, especially for historic documents, in places that are not as affluent or digitally oriented. SusunW (talk) 15:03, 12 January 2022 (UTC)

[Template:Cite encyclopedia] put the chapter along with the name of the entry

I am trying to add the following:

Melton, J. Gordon (2009). "Chapter 2 - Western Liturgical Family, Part I: The Western Catholic Tradition". Hugh George de Willmott Newman. Melton's encyclopedia of American religions (8th ed.). Detroit: Gale Cengage Learning. p. 84. ISBN 978-0-7876-9696-2.

However, the name of the entry ("Hugh George de Willmott Newman") is italicised. The only I can get the entry name to be between quotation marks is by removing or emptying the chapter parameter. Is there any way I could have something like ('Chapter 2 - Western Liturgical Family, Part I: The Western Catholic Tradition') to be displayed so that I can indicate clearly the name of the entry while also indicating the name of the chapter the entry is in? Veverve (talk) 16:05, 24 December 2021 (UTC)

@Veverve: Unfortunately, the Citation Lua code only recognizes the a pair of "location in reference" and "reference name" in {{Cite xx}} templates. For books, it's (|chapter=, |title=); encyclopedias, journals, and magazines, it's (|title=, |encyclopedia=), (|title=, |journal=), and (|title=, |magazine=).

To do what you want, you'll need to put the chapter after the template. But really, given entry title of "Hugh George de Willmott Newman" and the page number, the chapter isn't really necessary; it's just a little bit over-specificity.  — sbb (talk) 19:23, 24 December 2021 (UTC)

@Sbb: This encyclopedia is divided into chapters, with each chapter having its own alphabetical order from A to Z. Since this encyclopedia is updated every few years, I want to insure the reader will be able to find the entry I pointed to in the most recent editions in the future (and I am already one edition behind the current one). Do you think it is possible to edit the template so that it can do what I want? Veverve (talk) 19:47, 24 December 2021 (UTC)

How peculiar. There are three entries on pp. 84, 85: Joseph Rene Vilatte → Hugh George de Willmott Newman → Brazilian Catholic Apostolic Church. I presume that there is some rationale for that organization?

Regardless, you can spoof {{cite encyclopedia}} to some extent:

{{Cite encyclopedia |entry=Hugh George de Willmott Newman |encyclopedia=Melton's encyclopedia of American religions |publisher=Gale Cengage Learning |url=http://archive.org/details/meltonsencyclope0008melt |last=Melton |first=J. Gordon |date=2009 |publication-place=Detroit |edition=8th |page=84 |isbn=978-0-7876-9696-2 |title=''Chapter 2 - Western Liturgical Family, Part I: The Western Catholic Tradition''}}

Melton, J. Gordon (2009). "Hugh George de Willmott Newman". Chapter 2 - Western Liturgical Family, Part I: The Western Catholic Tradition. Melton's encyclopedia of American religions (8th ed.). Detroit: Gale Cengage Learning. p. 84. ISBN 978-0-7876-9696-2.

Alas, the important bit that you are citing (the value in |entry=) will not be available to readers who consume cs1|2 citations by way of the metadata. Do not add quote marks to |title= because they are not stripped from the value when it is made part of the metadata.

I want to insure the reader will be able to find the entry I pointed to in the most recent editions in the future. This is contrary to WP:SAYWHERE. We really can't future-proof citations. In some future version of Melton's, the entry may no-longer support the text at en.wiki; the chapter organization may change; other stuff may change... but the 2009 edition does support the en.wiki text so that is how the citation should be constructed. Were it me, I would write:

{{Cite encyclopedia |entry=Hugh George de Willmott Newman |entry-url=https://archive.org/details/meltonsencyclope0008melt/page/84/mode/2up |encyclopedia=Melton's encyclopedia of American religions |publisher=Gale Cengage Learning |url=http://archive.org/details/meltonsencyclope0008melt |url-access=registration |last=Melton |first=J. Gordon |date=2009 |location=Detroit |edition=8th |page=84 |isbn=978-0-7876-9696-2}}

Melton, J. Gordon (2009). "Hugh George de Willmott Newman". Melton's encyclopedia of American religions (8th ed.). Detroit: Gale Cengage Learning. p. 84. ISBN 978-0-7876-9696-2.

This points readers to the correct source in both the visual and metadata renderings.

—Trappist the monk (talk) 20:59, 24 December 2021 (UTC)

There is a third level of titling, between individual entries and the whole encyclopedia, given by |department=:

• {{Cite encyclopedia |entry=Hugh George de Willmott Newman |encyclopedia=Melton's encyclopedia of American religions |publisher=Gale Cengage Learning |url=http://archive.org/details/meltonsencyclope0008melt |last=Melton |first=J. Gordon |date=2009 |publication-place=Detroit |edition=8th |page=84 |isbn=978-0-7876-9696-2 |department=Chapter 2 - Western Liturgical Family, Part I: The Western Catholic Tradition}}
• Melton, J. Gordon (2009). "Hugh George de Willmott Newman". Melton's encyclopedia of American religions. Chapter 2 - Western Liturgical Family, Part I: The Western Catholic Tradition (8th ed.). Detroit: Gale Cengage Learning. p. 84. ISBN 978-0-7876-9696-2.

That may be a better choice in this instance. I think the part that gets left out of the computer-readable metadata is the department but does anyone actually use that format? —David Eppstein (talk) 21:22, 24 December 2021 (UTC)

@David Eppstein: Interesting. I guess I ignored that |department= was valid for {{Cite encyclopedia}}. I only use |department= with {{Cite newspaper}} or {{Cite magazine}} where the entry is something like an obituary, public notices, standing "letter from the editor", etc.  — sbb (talk) 23:40, 24 December 2021 (UTC)

Yes, I wasn't sure it would work until I tried it in this example. I usually only use it for journal, magazine, or newspaper citations. —David Eppstein (talk) 23:42, 24 December 2021 (UTC)

Yet another alternative: |at=

{{Cite encyclopedia |chapter=Chapter 2 - Western Liturgical Family, Part I: The Western Catholic Tradition |encyclopedia=Melton's encyclopedia of American religions |publisher=Gale Cengage Learning |url=http://archive.org/details/meltonsencyclope0008melt |url-access=registration |last=Melton |first=J. Gordon |date=2009 |location=Detroit |edition=8th |at=p. [https://archive.org/details/meltonsencyclope0008melt/page/84/mode/2up 84: Hugh George de Willmott Newman] |isbn=978-0-7876-9696-2}}

Melton, J. Gordon (2009). "Chapter 2 - Western Liturgical Family, Part I: The Western Catholic Tradition". Melton's encyclopedia of American religions (8th ed.). Detroit: Gale Cengage Learning. p. 84: Hugh George de Willmott Newman. ISBN 978-0-7876-9696-2.

All of the entry, chapter, and title are in the citation's metadata. Of course, the entry information is at the wrong end of the visual rendering...

—Trappist the monk (talk) 23:49, 24 December 2021 (UTC)

@Trappist the monk and David Eppstein: I may have mistaken in Newman's case a paragraph title with an entry, which is why there is not alphabetical order in the items Trappist the monk listed. Still, your help will be valuable to me for other articles; you can see an example at Christ Catholic Church (Pruter). I used David Eppstein's version. Veverve (talk) 23:53, 24 December 2021 (UTC)

Internationalisation need

Module:Citation/CS1/Configuration has local date information along with English language date information. This is a creating problem for Telugu Wikipedia users, as everytime this module is refreshed from enwikipedia, as part of import of Templates that use this module, the Telugu language information is getting overwritten. Then a manual update of Telugu language language dates is required to avoid check date errors being shown for Telugu dates. I request the maintainers to internationalise that portion. Arjunaraoc (talk) 00:30, 25 December 2021 (UTC)

By doing what, exactly? We will not add all of every language's possible date names as that would be undue burden of page processing on English Wikipedia. I think it's fairly reasonable to expect users refreshing the configuration page to check what they are doing and refresh only the relevant parts. Izno (talk) 05:03, 25 December 2021 (UTC)

@Izno, I am hoping that the local dates information can be stored in a subpage with language suffix. If such a page is not available, the default English dates can be used as local dates as well. On Telugu wikipedia, there are many admins who refresh the templates for their need, but do not have knowledge of template code or often forget to update the local date info. Hope that helps. Arjunaraoc (talk) 12:50, 29 December 2021 (UTC)

This is on my TODO list for after the next update.

—Trappist the monk (talk) 11:40, 25 December 2021 (UTC)

@Trappist the monk, Glad to know that it is in your todo list. Thanks. Arjunaraoc (talk) 12:51, 29 December 2021 (UTC)

addition of 'quote-p' and 'quote-pp' to citiation templates

Can someone please add 'quote-p' and 'quote-pp' as aliases of 'quote-page' and 'quote-pages' respectively? -- PK2 (talk) 05:06, 29 December 2021 (UTC)

WTF is this? (templatestyles stripmarker in title= at position XX)

Tang, Jian; Oka, Takeshi (1999). "Infrared spectroscopy of H₃O⁺: the v₁ fundamental band". Journal of Molecular Spectroscopy. 196 (1): 120–130. Bibcode:1999JMoSp.196..120T. doi:10.1006/jmsp.1999.7844. PMID 10361062. {{cite journal}}: templatestyles stripmarker in |title= at position 26 (help)

There's no issue in this citation. That error message needs to be suppressed. Headbomb {t · c · p · b} 18:23, 29 December 2021 (UTC)

It's old news that you shouldn't use templates in certain parameters. This is an outcome of someone's deliberate decision to do so.

See also a recent-enough archive that you can go looking for a discussion on a related topic. Izno (talk) 18:56, 29 December 2021 (UTC)

"you shouldn't use templates in certain parameters" there's never been any such proscriptions on template use anywhere. Case in point

Smith, J. (1999). "Fake title NO³⁻
". Journal of Stuff. 1 (2): 3–4.

Headbomb {t · c · p · b} 19:30, 29 December 2021 (UTC)

It has been repeated here ad nauseum. I doubt you have somehow missed those discussions. Izno (talk) 19:53, 29 December 2021 (UTC)

Something false repeated many times doesn't suddenly become true. Headbomb {t · c · p · b} 20:26, 29 December 2021 (UTC)

{{Su}} has been marked with {{COinS safe|n}} since this edit 18 January 2015. {{COinS safe}} has been around since 18 May 2012‎ (Special:Permalink/493117278). The COinS section of the cs1|2 template documentation was created 12 January 2012‎ (Special:Permalink/470891851) and incorporated at this edit (11 January 2012). Yeah, the timing seems a little odd... Still, we have discouraged the use of templates in cs1|2 parameters for nearly a decade now.

—Trappist the monk (talk) 21:15, 29 December 2021 (UTC)

@Headbomb: How about replacing {{H3O+}} with <sub>...</sub> and <sup>...</sup> tags, like this:

Tang, Jian; Oka, Takeshi (1999). "Infrared spectroscopy of H₃O⁺: the v₁ fundamental band". Journal of Molecular Spectroscopy. 196 (1): 120–130. Bibcode:1999JMoSp.196..120T. doi:10.1006/jmsp.1999.7844. PMID 10361062.

Happy editing! GoingBatty (talk) 19:26, 29 December 2021 (UTC)

The point is that this shouldn't throw an error at all, not that you can half ass a workaround. Headbomb {t · c · p · b} 19:30, 29 December 2021 (UTC)

Yup, the old issue of the inferior coins scheme used here. As suggested, such errors (that have nothing to do with citations) should be suppressed. Instead, perhaps whoever maintains coins could be auto-pinged any time this happens. 65.88.88.91 (talk) 19:32, 29 December 2021 (UTC)

This wouldn't be Wikipedia without additional confusion: some templates do work anywhere including |title=. One common example is {{en dash}}. For more, you can turn to the imaginary list documenting such templates that do not blame the victims. 65.88.88.91 (talk) 19:36, 29 December 2021 (UTC)

This seems to be related to <chem></chem> tags. Headbomb {t · c · p · b} 19:43, 29 December 2021 (UTC)

This should bypass the error. Headbomb {t · c · p · b} 19:49, 29 December 2021 (UTC)

Current

• Tang, Jian; Oka, Takeshi (1999). "Infrared spectroscopy of H₃O⁺: the v₁ fundamental band". Journal of Molecular Spectroscopy. 196 (1): 120–130. Bibcode:1999JMoSp.196..120T. doi:10.1006/jmsp.1999.7844. PMID 10361062. {{cite journal}}: templatestyles stripmarker in |title= at position 26 (help)

Sandbox

• Tang, Jian; Oka, Takeshi (1999). "Infrared spectroscopy of H₃O⁺: the v₁ fundamental band". Journal of Molecular Spectroscopy. 196 (1): 120–130. Bibcode:1999JMoSp.196..120T. doi:10.1006/jmsp.1999.7844. PMID 10361062. {{cite journal}}: templatestyles stripmarker in |title= at position 26 (help)

Headbomb {t · c · p · b} 19:50, 29 December 2021 (UTC)

If there is no list of CS1-safe templates, is there a list of coins-safe tags (if any tags are indeed compatible)? Until such issues are permanently resolved, template editors could consult that list and insert a notice re: usage in CS1 depending on the tags' existence in template code. 71.245.250.98 (talk) 21:07, 29 December 2021 (UTC)

I have reverted that edit because it doesn't do what you think it does. The sandbox rendering above, appeared to work because the {{H3O+}} template is not in the sandbox template (it was replaced with H<sub>3</sub>O<sup>+</sup>) so, of course, the 'fix' appeared to work.

The <chem>...</chem> markup is like <math>...</math> markup: the rendering that is visible is an image:

${\ce {H3O+}}$ ← <chem>H3O+</chem>

We might want to handle <chem>...</chem> markup in the same way that we handle <math>...</math> markup – some sort of error message in the metadata instead of a non-sensical stripmarker. Discussion is appropriate, I think.

—Trappist the monk (talk) 21:15, 29 December 2021 (UTC)

If <chem>...</chem> is like <math>...</math>, then it should be handled like math. The revert is nonsensical even if it didn't fix the above issue. Headbomb {t · c · p · b} 21:31, 29 December 2021 (UTC)

@Headbomb: It appears that {{H3O+}} uses {{chem}} (not <chem>...</chem>), and Template:Chem's documentation includes {{COinS safe|n}}, with a note stating that <chem>...</chem> is an alternative to {{chem}}. GoingBatty (talk) 23:21, 29 December 2021 (UTC)

Looking cursorily at the source code of {{chem}} and the subtemplates {{chem/atom}} and {{chem/link}}, as well as module:su nothing stands out as particularly offensive html-tag-wise. One wonders what trips COinS. Ten+ years in a row. 69.203.140.37 (talk) 13:59, 30 December 2021 (UTC)

Incidentally, none of the versions of the citation above are formatted correctly. The $\nu _{1}$ part is a mathematical equation and should be formatted as mathematics, either as <math>\nu_1</math> or as {{math|''ν''₁}}. Additionally, I'm pretty sure that letter in it is a Greek nu, not a Latin vee. The <math>\nu_1</math> should always work in titles but will stay black even when the title is linked by a url or free doi. The {{math|''ν''₁}} will look correct when linked and I think looks nicer here but may run into similar stripmarker issues.

• {{cite journal | title = Infrared spectroscopy of {{H3O+}}: the <math>\nu_1</math> fundamental band | first1 = Jian | last1 = Tang |first2 = Takeshi | last2 = Oka | journal = [[Journal of Molecular Spectroscopy]] | volume = 196 | pages = 120–130 | year = 1999 | doi = 10.1006/jmsp.1999.7844 | pmid = 10361062 | issue = 1| bibcode = 1999JMoSp.196..120T }}
• Tang, Jian; Oka, Takeshi (1999). "Infrared spectroscopy of H₃O⁺: the $\nu _{1}$ fundamental band". Journal of Molecular Spectroscopy. 196 (1): 120–130. Bibcode:1999JMoSp.196..120T. doi:10.1006/jmsp.1999.7844. PMID 10361062. {{cite journal}}: templatestyles stripmarker in |title= at position 26 (help)
• {{cite journal | title = Infrared spectroscopy of {{H3O+}}: the {{math|''ν''₁}} fundamental band | first1 = Jian | last1 = Tang |first2 = Takeshi | last2 = Oka | journal = [[Journal of Molecular Spectroscopy]] | volume = 196 | pages = 120–130 | year = 1999 | doi = 10.1006/jmsp.1999.7844 | pmid = 10361062 | issue = 1| bibcode = 1999JMoSp.196..120T }}
• Tang, Jian; Oka, Takeshi (1999). "Infrared spectroscopy of H₃O⁺: the ν₁ fundamental band". Journal of Molecular Spectroscopy. 196 (1): 120–130. Bibcode:1999JMoSp.196..120T. doi:10.1006/jmsp.1999.7844. PMID 10361062. {{cite journal}}: templatestyles stripmarker in |title= at position 26 (help)

—David Eppstein (talk) 18:00, 30 December 2021 (UTC)

Referencing equations in physics books

At Carnot cycle we want to reference https://www.gutenberg.org/files/50880/50880-pdf.pdf, specifically we want equations 39, 40 and 65 in sections §90 and §137. I'm not sure of the best way to do this.  Stepho  talk  00:59, 30 December 2021 (UTC)

Write something like <ref>See equations 39, 46 and 65 in {{cite ...}}</ref> or <ref>{{cite ...|at=Eqs 39, 40, 65}}</ref> Headbomb {t · c · p · b} 01:08, 30 December 2021 (UTC)

Although there's nothing wrong with doing it those ways, you could also use |contribution=Equations 39, 40, and 65 in Sections 90 and 137 within the citation template. Doing it that way has the advantage that you can still specify pages as a separate parameter rather than needing to format the page numbers as part of the |at= parameter, and that you can provide a direct link to the starting page of the contribution (if you're using a source like archive.org or Google books that provides such links) in the |contribution-url= parameter. —David Eppstein (talk) 01:59, 30 December 2021 (UTC)

Thanks. I went with |contribution=equations 39, 40 and 65 in sections §90 & §137, although I was tempted to use |section=equations 39, 40 and 65 in sections §90 & §137. I was able to add |page=.  Stepho  talk  11:18, 30 December 2021 (UTC)

Check pmc=value

"If the value is correct and larger than the currently configured limit of 8700000, please report this at Help talk:Citation Style 1, so that the limit can be updated."

Molecular mechanisms for understanding the association between TMPRSS2 and beta coronaviruses SARS-CoV-2, SARS-CoV and MERS-CoV infection: scoping review has PMC=8709906

-- Ben Best:Talk 21:32, 30 December 2021 (UTC)

access-date for archived pages?

For citations that are archived from the very start (i.e. cited only after the original website has lapsed), what should access-date in Template:Cite web contain: the last time the writer successfully accessed the original URL, or the last time the writer successfully accessed the archived URL? The documentation defines the parameter as “[t]he full date when the original URL was accessed” – so should the parameter perhaps not be present at all in that case? Obskyr (talk) 04:59, 4 January 2022 (UTC)

I personally remove access-date when an archive is provided. Otherwise, it should be the date the original URL was last accessed. Izno (talk) 06:34, 4 January 2022 (UTC)

I agree with Izno. I think having an access date in these circumstances is pointless and just causes confusion. -- Alarics (talk) 10:17, 4 January 2022 (UTC)

Alright, but wait a minute. I see a chart peak on Billboard's ever-changing site and add a cite to the "That Great Song" page. The weekly chart position changes, up or down, maybe, and then the page is archived. Billboard revamps their stupid site again and the page, as it once was, is gone forever, except at the archive (and maybe some other web site, and maybe in a paper magazine), although possibly with different chart placement. Is it really wrong to say, "I saw That Great Song charted at No. 9 on so-and-so date at this URL"? Isn't that we're supposed to do? — JohnFromPinckney (talk / edits) 20:18, 4 January 2022 (UTC)

There are 2 dates at play here, if I understand you correctly. The event date and the access date of the archive. The event date is a week, as Billboard charts by week, and obviously is a dynamic date - even if the song charts at the same position for many weeks, this will eventually change. Your access date sometime later will most likely be the date you access an archive of the event. The archive is static, and therefore archive access dates are thought to be unimportant. However the archive date is important, as it relates the archive to the event you want to cite. |archive-date= will suffice in this case, obviously as long as it is later than the event date. 65.88.88.57 (talk) 20:37, 4 January 2022 (UTC)

No. |access-date= is the date that the web page at the original URL was used to verify the content in our article, and |date= is the date that this original web page was published; |archive-date= is the date that Wayback Machine (or whoever) made an archive copy of the web page concerned. --Redrose64 🌹 (talk) 21:01, 4 January 2022 (UTC)

Depends on the archive. Crawlers such as the Wayback Machine store snapshots for some time (it is not entirely clear to me how the inventory of snapshots is handled, but it is not unusual for snapshots to disappear). The archived content (ie snapshots) of a web page @ the Wayback Machine is not static; only the content of the snapshots is. The number of snapshots per archive is likely changing with every crawl of the page. When a citation calls an archive in the Wayback Machine, it actually calls a particular snapshot that includes the pertinent information. Therefore |archive-date= is important. There may be several snapshots on that date, but CS1 currently does not provide snapshot resolution. But the original question was about the access date of the archive. To answer this properly, we have to know how each service handles its archives. Empirically, most people accept archives as static, regardless of the service provider. 65.88.88.201 (talk) 14:50, 5 January 2022 (UTC)

Adding Authors Wikidata ID

Given that a noted journalist or academic that authored the text thats being referenced will themselves be able to referenced, could we store their wikidata id (as well the first and surname) which in turn stores their various academic ID's (such as ORCID) and popularist ID's (such as twiter), perhaps linking to reasonator to make it readable? It would also solve the the variation in their names, for example they may use a formal version of their name in academic papers and casual for something in the popular press as well varations whether they use "special characters" in their name

The same could be used for the publicaton

So for example

Pöhls, Jan-Hendrik. "A new approach finds materials that can turn waste heat into electricity". The Conversation. Retrieved 2022-01-05.

Would become something like

Pöhls, Jan-Hendrik. "A new approach finds materials that can turn waste heat into electricity". The Conversation. Retrieved 2022-01-05.

Back ache (talk) 15:05, 5 January 2022 (UTC)

@Back ache: Instead of asking users to add even more data in references, I'd suggest that we could use the existing |author-link= and |work= parameters to provide links to articles about the authors and publications, create redirects for variations in their names, and add ID links in the External links section of those articles. Happy editing! GoingBatty (talk) 16:00, 5 January 2022 (UTC)

That could work, in my earlier example then I'd add author-link=d:Q104888284 Back ache (talk) 23:09, 5 January 2022 (UTC)

You will want to review previous discussions (and there may be more where the phrase 'qid' is not uttered). Izno (talk) 19:47, 5 January 2022 (UTC)

And the obvious search for Wikidata. Izno (talk) 19:48, 5 January 2022 (UTC)

Instead of adding ID links individually to articles, they should be handled through adding them to Wikidata and using the {{Authority control}} template on the articles. But in references, I tend to think that only authors with existing Wikipedia articles should be author-linked; we shouldn't instead link to Wikidata items for authors without articles. Also, I tend to think your proposed link style violates Wikipedia:Manual of Style/Accessibility#Links. —David Eppstein (talk) 20:25, 5 January 2022 (UTC)

Whilst there will always be a place for narrative-first (wikipedia), I also like to assemble facts (wikidata) first and let that tell a story that can be made into an artical, you could argue sufficient good-quality links tell a story in their right and with things like resonator and the semantic things coming that data-only entries will continue to increase Back ache (talk) 21:51, 5 January 2022 (UTC)

It is probably best not to consider citations "narrative". They are tools meant to ensure the actual narrative is factual. This is a vital undertaking that should be focused, uncomplicated, understandable and simply implemented. As was suggested above, overhead should be kept to a minimum. As you will see from the previous related discussions, there are also several as yet unanswered questions about Wikidata in general. 68.173.76.118 (talk) 02:23, 6 January 2022 (UTC)

OPPOSE Wikidata author IDs do not belong in citations. This is complete clutter. The only thing that might belong in a citation is the Wikidata QID for the article/chapter/book being cited, if the article/chapter/book has a corresponding Wikidata entry. Headbomb {t · c · p · b} 12:57, 6 January 2022 (UTC)

Template-protected edit request on 6 January 2022

Could somebody add "archive-link" (correct to " archive-url)? Thanks in advance. Leomk0403 (Don't shout here, Shout here!) 09:25, 6 January 2022 (UTC)

If you are asking for cs1|2 to support |archive-link= as an alias to |archive-url=, I would say that we ought not do that. link can be taken to mean more than just a url (a wikilink is a link...) Editors know and understand |archive-url= so I see no benefit to adding |archive-link= as an alias.

—Trappist the monk (talk) 14:12, 6 January 2022 (UTC)

I have added this to the Suggestions page, as requested. Someone who uses archive-link will receive a red error message with a helpful suggestion to use archive-url instead. – Jonesey95 (talk) 20:19, 6 January 2022 (UTC)

Error messages for non-usual date formats

The guidance of the CS1 help page states that 'Sources are at liberty to use other ways of expressing dates, such as "spring/summer" or a date in a religious calendar; editors should report the date as expressed by the source.' However, the templates as implemented do not appear to respect this, as the date errors are not suppressable. How should this be addressed? --Paul_012 (talk) 11:25, 6 January 2022 (UTC)

An example would be helpful. 71.247.146.98 (talk) 12:48, 6 January 2022 (UTC)

Continue reading, Paul_012. In cases where the date as expressed in the source is not compatible with the template software, the citation should be created without using a template. – Jonesey95 (talk) 13:25, 6 January 2022 (UTC)

Huh. Not sure how I missed that. Does anyone know how this advice originated? It seems rather suboptimal to abandon the templates altogether just because the dates don't fit. --Paul_012 (talk) 14:01, 6 January 2022 (UTC)

This edit.

—Trappist the monk (talk) 14:09, 6 January 2022 (UTC)

On the contrary, templates optimize common cases, sometimes with allowance for very few exceptions. The important item here is the citation, not how it is produced. 71.247.146.98 (talk) 14:13, 6 January 2022 (UTC)

Memento Project

This is not directly related to CS1, but it is something to be aware of. This project ("About Memento" link) is basically a meta-archive whose protocol (RFC 7089) includes logic to automatically access the optimal version of an archived page among different archives. Perhaps more importantly for CS1, there is no need for |archive-url=; the built-in logic in the Memento protocol parses the original URL and determines its status, and whether an archived version exists, which is then served transparently to the user. The current drawback is that the Memento protocol has to be called via a browser extension. It is conceivable that it may become a built-in feature. I am not aware of any Wikipedia template/script that applies this protocol currently. 65.88.88.62 (talk) 16:06, 11 January 2022 (UTC)

What exactly is the issue with TemplateStyles in arguments?

As discussed above, TemplateStyles tags (or rather their strip markers) flag up as an error: "Wikipedia". {{cite web}}: templatestyles stripmarker in |title= at position 1 (help). What exactly is the purpose of this? It makes it much more difficult for template editors to add TemplateStyles to any inline template, as there's a risk that someone has used that template in a citation and citations will then start issuing errors.

If it's a risk of corrupting COinS, why not just silently strip it from machine-readable parts of the output? By their nature the TemplateStyles strip markers only need to appear once to work. If even having it inside the displayed title (or wherever) is problematic, then it could be moved from there to before the rest of the citation. User:GKFX^talk 19:29, 11 January 2022 (UTC)

This is your example citation:

{{cite web|title={{smallcaps|Wikipedia}}|url=https://en.wikipedia.org}}

What cs1|2 gets (after {{smallcaps}} has been processed is this:

?'"`UNIQ--templatestyles-00000051-QINU`"'?<span class="smallcaps">Wikipedia</span>

(the '?' on either end of the stripmarker is the delete character, U+007F)

Of all of that, the only part that is the title, the only part that belongs in |title=, is 'Wikipedia'; the rest is extraneous junk that does not belong in |title=.

The strip marker detector is a relatively easy way to identify corrupted cs1|2 template parameter values. The detector looks for all flavors of strip marker. When templatestyles strip markers are found in |id= or |quote=, they are ignored because those parameter values are not made part of the citation's metadata. Math strip markers are removed from the metadata and replaced with 'MATH RENDER ERROR' message because Scribunto no longer allows modules to fetch the content of math strip markers. A math strip marker remains in the rendered parameter value so that MediaWiki can replace it with an image of the equation. cs1|2 can fetch the content of nowiki strip markers so it does so when creating metadata. All other strip markers are considered erroneous.

Silently removing strip makers is easy. It is more difficult and time consuming to remove the rest of the junk that templates can add to a parameter value. We had a relatively long and ultimately unproductive discussion about unnecessary html markup at Help talk:Citation Style 1/Archive 80 § HTML markup.

—Trappist the monk (talk) 23:12, 11 January 2022 (UTC)

Thank you, that was informative. I'm mostly interested in {{chem2}}; what did you think of the suggestion at Special:Diff/915129724/1046369074? If the citation templates accepted some indicator of the correct plain-text representation, it would be fairly straightforward to build a plain text representation within chem2 at the same time and put that out in an HTML attribute or equivalent. The input to chem2 is semi-readable to a chemist, which I think you asked, but not exactly presentable, so a specific plain text output would be necessary. Ideally that would also silence the strip marker error. User:GKFX^talk 23:31, 11 January 2022 (UTC)

I think that the markup-in-span scheme may be an idea worth pursuing. I think that it should be limited to only those kinds of templates that 'translate' markup-in-plain-text to some sort of standard presentation (as LaTeX or TeX is used for math). I don't know if semi-readable is good enough. I asked that question because en.wiki should never be in the business of making-up 'standards' for anything especially when there are already extant standards in the outer world. We have <chem>...</chem> tags so presumably what goes inside those tags (apparently some form of LaTeX) is standardized markup. That is the markup-in-plain-text that should be used by {{chem2}} and supplied in the markup-in-span scheme so that readers who consume a cs1|2 template via its metadata can read and understand (because it is standardized) the markup that makes an equation in |title=.

I also think that the markup-in-span proposal is something that deserves (requires?) buy-in from more than just you and me. Is {{math}} one such template that might use the scheme? And the templates that are used in it? What about nesting of one (or more) markup-in-span template inside another markup-in-span template? Are there other templates that might be suited to the scheme? If not, then perhaps it is not worth the effort to create and maintain the markup-in-span scheme. Styling templates like {{smallcaps}} have no reason to be included in the scheme because styling of cs1|2 citations is the responsibility of cs1|2.

In the markup-in-span scheme the span's class= attribute would be sufficient to turn off the stripmarker error.

—Trappist the monk (talk) 14:38, 12 January 2022 (UTC)

I'm unconvinced that the <chem> tag's markup (LaTeX with the mhchem library) is that important of a standard; it's just one library. The notation it uses like HC#CH → ${\ce {HC#CH}}$ is not something I encountered at university, and I wouldn't know that # meant triple bond if I hadn't read it in Wikipedia's documentation. It would be far better to output "HC≡CH" in that case, which is plain text and a correct notation. I think the question of what the exact plain text representation would be is one for WikiProject Chemistry; once you get to stuff like ¹⁶O2−2 it's less obvious; something LaTeX-ish in that case seems appropriate. For {{math}}, it looks like what matters is the templates inside math (e.g. {{radic}}) as math seems to just apply formatting. Nesting could be awkward, but if radic output "√(" (as the hidden plain text), then its first parameter, then ")", the citation template would just have to concatenate all the hidden plain text it saw and nesting would have worked successfully.

In terms of the HTML representation of this, I would have thought an attribute data-wiki-plaintext="HC≡CH" would be more appropriate than putting it as document text in a span and then having to hide the span with CSS. User:GKFX^talk 21:39, 12 January 2022 (UTC)

If you're going to use "something LaTeX" for some of these notations, you need to use it for all of them, because of Wikimedia's bad choices to render LaTeX in a way that is visually incompatible with other formatting styles. —David Eppstein (talk) 21:53, 12 January 2022 (UTC)

That's not what I meant, I meant that if I cited a paper called "Properties of O2−2", it might appear as "Properties of O2^{2-}" in the COinS metadata, as you can't output "O22-". On the article the appearance of chem2 would not change from its current HTML. User:GKFX^talk 21:58, 12 January 2022 (UTC)

I have no experience writing chemistry equations. But, I find it very hard to believe that the chemistry world doesn't have a standardized way of creating chemistry equations. Surely the researcher who submitted the paper called "Properties of O2−2" to the journal for publication, wrote that, and other, equations using a method that guaranteed that the publisher would typeset it correctly. After all, that is all that {{chem2}} and like templates are doing here. Some sort of mechanism is required to place all of the various parts of the equation in their proper positions. Is that something that generic word processors do these days? If not then what is the typesetting standard used by industry journals? Whatever that standard is, that should be the standard used at en.wiki. If there is no such standard then, I suppose, en.wiki can do whatever it wants when it comes to typesetting chemistry equations. You don't like LaTeX with the mhchem library, suggest a better standard.

Please don't chastise me for something that I did not write. You asked for my comments about this {{Chem2/sandbox}}. That sandbox uses the class= attribute. Were we to implement something like that, regardless of whether it uses class= or data-whatever=, cs1|2 would remove that attribute thing from its input parameter before rendering so no need to hide the span with CSS.

—Trappist the monk (talk) 23:37, 12 January 2022 (UTC)

The major issue is not how to render chemical formulae, but rather how to generate correct metadata. That is an issue whether you are using, e.g., LaTeX, MarkDown, wiki. What about separate parameters for display title and reference titles? --Shmuel (Seymour J.) Metz Username:Chatul (talk) 15:33, 13 January 2022 (UTC)

Correct, and as said below, this is not a CS1/2 issue. All such discussions should be redirected at the appropriate forum for metadata in Wikipedia, not here. This comes up over & over again, so there is no confidence that this is going to be resolved any time soon. Just one more time-wasting burden for editors who want to use CS1/2. The discussion above is proof enough - instead of removing the error, all kinds of unnecessary contortions are proposed to accommodate it. 65.88.88.47 (talk) 16:57, 14 January 2022 (UTC)

It is more difficult and time consuming to remove the rest of the junk that templates can add to a parameter value. This is not pertinent. Editors do not need templates to add junk to a parameter value, and the fact that it is indeed difficult to remove has nothing to do with this discussion. The main, longstanding issue is that COinS may interfere with legitimate citation editing. This is not new. A thorough search will discover complaints going back at least 12+ years, or ever since the introduction of the scheme in the defunct {{citation/core}}. Btw, the discussion on adding the metadata scheme (if memory serves, prompted by a request from Zotero-using editors) was minimal, as I recall. The problem is that COinS is based on OpenURL, which basically adds specific-content-location additions to a regular http address. Extraneous http artifacts will obviously generate unwanted results. It is incumbent on Wikipedia COinS to find a workable solution to that problem instead of limiting legitimate usage of templates that make editors' work easier. I've no idea what constituency of Wikipedia will agree to limit editors in favor of an external scheme, but it may be time to find out. Keeping also in mind that not all templates will generate COinS-based errors, which makes blanket statements about template use in citations somewhat arbitrary. Until Wikipedia COinS behaves, the suggestion of adding a switch to turn metadata off is an excellent one, and it should be implemented. 65.88.88.91 (talk) 21:10, 12 January 2022 (UTC)

[Template:Cite web] Website parameter being misused?

The {{Cite web}} docs clearly state that the website parameter should display the name of the site, and the examples back that up. Though for some reason, on lots and lots of articles that I see, it's used to display the domain name. For example, where I would think someone would put |website=[[Microsoft]] on a page on Microsoft's website, I often see |website=www.microsoft.com or something similar, instead. Have I misunderstood what the examples, docs, and templatedata say? Or is everyone using this parameter incorrectly. Thanks, ― Levi_OP^Talk 16:37, 13 January 2022 (UTC)

You have not misunderstood. Unfortunately, the doc was not as clear as it should have been in the past, and it is not unusual for the incorrect format to appear in older citations. Also, perhaps editors do not pay proper attention to the current documentation. 65.88.88.68 (talk) 16:45, 13 January 2022 (UTC)

Sooo should I fix this on any articles I see it on? This seems like a pretty big task that some bot could do instead considering just how many articles have this issue. ― Levi_OP^Talk 16:50, 13 January 2022 (UTC)

It's often bots and AWB that do it in the first place because they can easily determine via the URL but they can't know what the name should be. -- GreenC 16:54, 13 January 2022 (UTC)

(edit conflict) Microsoft is not a website, so you should not find |website=[[Microsoft]]. Those should be replaced with |publisher=[[Microsoft]]. |website=www.microsoft.com is fine, thought not really needed if you have |publisher=[[Microsoft]]. Headbomb {t · c · p · b} 16:55, 13 January 2022 (UTC)

Oh, thanks. I didn't know that this publisher option existed, so it would be better in this case. Although, it was just an example. I see it with other links where it is a website and not a company page or something of the sort. ― Levi_OP^Talk 17:16, 13 January 2022 (UTC)

In the case of websites, the publisher is the hosting entity, usually the domain-name owner. This applies when there is a different website creator/owner, as it happens with "free" websites offered by ad-supported web hosting services (actually such websites are usually sub-domains, but that's another story). However, the trade name or dba of the publisher is appropriate if different from the domain-name, so use |publisher=Microsoft rather than |publisher=microsoft.com. 65.88.88.75 (talk) 19:46, 13 January 2022 (UTC)

@GreenC: How would AWB "easily determine via the URL" and create a malformed reference? Maybe you were thinking of tools like Reflinks and/or reFill? GoingBatty (talk) 20:40, 13 January 2022 (UTC)

Well AWB by default can not, but users can create external scripts/programs that hook into AWB (Python, JS, PHP etc). Basically AWB handles the authentication, download and and upload of the page, and a user script processes the page to do whatever. Many bots run this way. Yes the user-initiated tools are also a factor. -- GreenC 20:45, 13 January 2022 (UTC)

|website=www.microsoft.com is not fine. It is the name of the host where the title (and maybe not the whole web site) can be found. It is different from the name of the web site, which in this case happens to be the same as the name of the publisher, Microsoft (although you could spell out the publisher as "Microsoft Corporation" if you were feeling pedantic, and the name of the site would still just be Microsoft). In general, my philosophy for sites where the publisher is more informative than the whole-site name is to use the "work" parameter to give some context to the specific title being cited — is it part of a larger thing, and which level of larger thing would tell readers the most useful information about the citation? For the same reason we neither use street numbers nor "Earth" to quickly describe the residence of someone, we should often neither use the most specific description of a larger work containing the title nor the most general description ("World Wide Web") to describe the work it comes from, but something intermediate, chosen to be informative. Maybe that's the title of a magazine or newspaper, rather than the department within that publication, or the group of publications ("Times-Mirror Newpapers") it comes from. In the case of [2], for instance, (one of the references in Microsoft), the publisher is "Microsoft", and it is indeed somewhere in Microsoft's vast web site, so it is not wrong to put |work=Microsoft, but a more informative and therefore better choice would be |work=Official Microsoft Blog. To complete the analogy, putting a domain name in the work parameter would be like writing that someone lives in the 02134 zip code — it is informative, roughly at the level a city name might be, but in a format primarily intended for conveying information unambiguously to computers, rather than for communicating to people. The text of an encyclopedia article (and a citation is part of that text) should be for communicating to people. —David Eppstein (talk) 23:00, 13 January 2022 (UTC)

Microsoft is not a website, it's a publisher. The name of the website, if anything is Microsoft.com. Headbomb {t · c · p · b} 04:12, 14 January 2022 (UTC)

Looking at first principles, why add a website/publisher: to facilitate verification. Sometimes no archives exist, so additional metadata is useful in finding a copy in databases such as LexisNexis and others. The more 'fancy' one gets with naming, the more likely a mismatch. For example if LN saved it as "Microsoft.com" then searching on "Official Microsoft Blog" might not find it. But searching on only "Microsoft" would find it. Thus it would probably be preferable to have simply |publisher=Microsoft - given V as the consideration concision and common name is important, similar to our article naming conventions, ease of finding things. The idea that precision is better for verification makes sense, unless other websites (LN) have a different idea what to call the website/publisher, which is probably a common problem because there are no standards. -- GreenC 05:49, 14 January 2022 (UTC)

Adding the Index Theologicus

Hello. I think it would be a good idea to be able to add the Index Theologicus identifier to the templates the way the DOI and the JSTOR identifier are shown. For example, "IxTheo 158777335X" would be displayed with IxTheo hyperlinked and an URL at 158777335X linking to here; so it would look like: "IxTheo 158777335X". This would be useful for articles which do not have a DOI or a JSTOR identifier such as the one used in my example. What do you think? Veverve (talk) 03:56, 14 January 2022 (UTC)

The templates support general parameters in general. If you would like to add this as an identifier for a given use, you may use |id=, possibly in combination with some formatting template (like exists today with {{doi}}). Izno (talk) 07:30, 14 January 2022 (UTC)