Help talk:Citation Style 1/Archive 18

This is an archive of past discussions about Help:Citation Style 1. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 15

Archive 16

Archive 17

Archive 18

Archive 19

Archive 20

→

Archive 25

Books can be uniquely identified by ISBN, and journals by DOI and other identifiers, which are included in citations. Distinct identification of authors, where names may change, be abbreviated or even misspelled, or variously translated, has been a problem, but is now being addressed by databases such as ORCID, which provide unique identifiers as well as comprehensive information on an author and his or her work. As such identifiers provide valuable bibliographic data I have developed a template to provide links to an author's databae entry.

The {{Authorid}} template, given the proper identfiers, generates links to an author's information in the form of one or more superscripted letters. Example: Josiah Carberry^o s. This can be used now in citation templates by appending it to the |first= parameter. However, that is not suitable, as 1) it pollutes the COINS data, and 2) the links fail when used with |author-link=.

I propose we have an |author-id= parameter (and the corresponding |authorn-id= parameters) to provide a place for author identifiers such as ORCID, ResearcherID, etc. Note: this is not to generate the links, nor to display the identifiers directly (which would overwhelm a citation), but to display links generated by Authorid (or similar templates) in the proper place following an author's name.

Comments? ~ J. Johnson (JJ) (talk) 20:46, 10 April 2016 (UTC)

I think that Pigsonthewing (talk · contribs) has already requested this. The relevant thread will be in the archives somewhere. --Redrose64 (talk) 22:39, 10 April 2016 (UTC)

My request was for ORCID iDs specifically. The most recent discussion was in April 2014. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:36, 11 April 2016 (UTC)

There's some discussion of the proposed changes on my talk page; in particular the new template needs to use {{abbr}} (or equivalent), for accessibility. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:36, 11 April 2016 (UTC).

Please keep in mind that what I am proposing here is to provide a place for author identifiers in the citation templates, however such identifiers are provided. ~ J. Johnson (JJ) (talk) 21:08, 11 April 2016 (UTC)

Are you suggesting the equivalent of |at= for authors? --Izno (talk) 23:11, 11 April 2016 (UTC)

I am not certain exactly what you mean, but possibly "yes". At least in regard of accepting free-form content. ~ J. Johnson (JJ) (talk) 01:29, 12 April 2016 (UTC)

I don't see it as Wikipedia's role to further clutter our editing window with ORCID data. The information is virtually pointless, and does no service to the reader. Headbomb {talk / contribs / physics / books} 19:36, 13 April 2016 (UTC)

ORCID iDs provide a number of services to our readers; not least enabling them to disambiguate authors with shared names; and to associate the work of a single author under different name-variants. Not what I would call pointless. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:48, 13 April 2016 (UTC)

And that's exactly what I meant by virtually pointless. See WP:DIRECTORY. Citations are there to provide verifiability. When I cite doi:10.3998/3336451.0004.203, it doesn't matter which of the many LA Davidson the article is referring to. If you're interested in that, follow the link and check the journal's website. Any journal that has implemented ORCIDs will have implemented DOIs well before that. Headbomb {talk / contribs / physics / books} 20:16, 13 April 2016 (UTC)

In an ideal world, that would certainly be the case. This is not an ideal world. You also assume that every work for which at least one author has an ORCID iD, there is a DOI. This is also not the case. WP:DIRECTORY does not prevent us from disambiguating our citations, and this proposal does not entail crating a directory. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:40, 13 April 2016 (UTC)

An essential requirement for sources used on WP is that they be reliable, and that is largely a function of the authors. Where an editor (or even a reader) is not familiar with a subject, it is of prime value to see an author's publication history, which is a principal purpose of databases like ORCID. It is also very useful for editors and readers alike in order to follow up on the most recent developments on a topic. Providing a link to such a database does not violate WP:NOTDIR.

If your edit window has gotten too cluttered perhaps the biggest improvement to be made is pulling all of the citation templates into their own section (replacing them with short cites).~ J. Johnson (JJ) (talk) 21:40, 13 April 2016 (UTC)

So is this proposal non-controversial? Or should the discussion be expanded, or possibly an RfC started? I suspect the real criterion for getting this implemented is simply getting Trappist on board. Trappist the monk, what do you think of this? ~ J. Johnson (JJ) (talk) 20:28, 17 April 2016 (UTC)

Isn't this topic rather closely related to the conversation at §Sources (or authors) with a Wikidata item, but no article? It doesn't appear to me that the issue is settled.

—Trappist the monk (talk) 14:02, 18 April 2016 (UTC)

Closely related in that both concepts are about providing information about authors, and both are premised on this being a good thing. But quite different approaches. The other concept is an alternative for the existing |author-link= (which is a wikilink), while my concept of |author-id= is for a space (platform?) that could accommodate a variety of features. Even a link to Wikidata that is independent of |author-link=. Wikidata links could even be accommodated in my {{authorid}} template. But whether Wikidata links are a good thing is entirely a matter of the other discussion, and not pertinent here. ~ J. Johnson (JJ) (talk) 20:40, 18 April 2016 (UTC)

The proposal is hardly non-controversial when of the two editors who have responded substantively, one opposes it and the other wants something different. I would also oppose it, on the grounds that it clutters the displayed citation with detail that is not part of the purpose of these templates, namely to allow readers to locate the work cited. Kanguole 21:32, 18 April 2016 (UTC)

As I have alrady said: the proposal here is to provide a place for author identifiers, not to implement author identifier links directly. The "something different" that Andy previously asked for was (as he has explained) for ORCID only, a possible alternative to {{authorid}}. That is in no way an objection to this proposal.

So in opposition are Headbomb (and I thought we had adequately dealt with his objections) and you. And your objection is that having one or more small, superscripted characters (like this ^o) is too much clutter, and "not part of the purpose of these templates". Well, what degree of "clutter" is unacceptable seems highly subjective, but it is no where near the "clutter" of isbns, dois, bibicodes, etc., which don't seem to bother you, and therefore I don't find your objection very credible. As to the purpose of citations, it is, as I have already said, an essential requirement that sources be reliable, and that is largely a function of authors. If you think that citations must be constrained to something much narrower, please give us some authority. ~ J. Johnson (JJ) (talk) 00:25, 19 April 2016 (UTC)

The difference with ISBNs, DOIs, etc, is that they identify and help to locate the bibliographical item being cited, which is the purpose of these templates. Reliability of sources is a vital part of editing, but that does not mean that we should expand citations with additional information to assist readers to make their own evaluation. The form suggested at {{authorid}}:

• Carberry, Josiah^o s; Bourbaki, Nicolas^v (2004). Mythical Maths. {{cite book}}: External link in |first1= and |first2= (help)CS1 maint: multiple names: authors list (link) CS1 maint: numeric names: authors list (link)

is both cluttered and obscure. It could be made less obscure, but that would increase the clutter. Kanguole 00:44, 21 April 2016 (UTC)

Again, this is just your personal, subjective opinion. If you really think citations should be so narrow (and you have yet to cite any authority why we must accept that) then perhaps you should work to remove |author-links=, as they also do nothing to identify or locate the source, and all those splashes of blue may (in some people's opinion) be just too distracting. And your objection to {{authorid}} seems tautologically impossible to satisfy: how can the result be both more informative and less intrusive? ~ J. Johnson (JJ) (talk) 02:22, 21 April 2016 (UTC)

It can't – that was my point. Kanguole 16:23, 21 April 2016 (UTC)

And my point is that this impossibility reflects on the absurdity of your arbitrarily chosen criterion. Apply it to dois, isbns, even author names, and all can be made to fail. It's just an indirect form of WP:IDONTLIKEIT. But we are getting off-topic, as the proposal here is create an |author-id= parameter. If you don't want to use that for the {{authorid}} template feel free not to. Or devise a better way of linking to author information. ~ J. Johnson (JJ) (talk) 18:58, 21 April 2016 (UTC)

In case it was unclear, I am opposed to adding parameters for additional author information, for the reasons given above. Kanguole 21:41, 21 April 2016 (UTC)

So we can move on regarding any objection to {{authorid}}. Regarding your main (?) point of a narrow interpretation of what citations should or should not do: do you have any authorities to support your view? ~ J. Johnson (JJ) (talk) 23:41, 21 April 2016 (UTC)

Identifying the work cited is what they do now. You and Andy have proposed to extend them to incorporate author identifiers. Clearly we have different opinions on whether that should be done. Kanguole 00:32, 22 April 2016 (UTC)

Clearly. But is your opinion alone sufficient to deny everyone else the benefit of author identifiers? ~ J. Johnson (JJ) (talk) 20:23, 22 April 2016 (UTC)

You are forgetting User:Headbomb. In any case, very few voices have been heard – perhaps you should open an RFC to get more opinions. Kanguole 22:53, 22 April 2016 (UTC)

I have not forgotten Headbomb, nor am I precluding a possible RfC, but at that this point my interest is in your view. Particularly if you any authority for it, or even a persuasive argument. ~ J. Johnson (JJ) (talk) 18:06, 23 April 2016 (UTC)

I have given my reasons above. You do not find them persuasive, but that is unsurprising. Kanguole 00:54, 24 April 2016 (UTC)

The reasons you gave (on 21:32, 18 April and 00:44, 21 April) are: 1) the {{authorid}} template is simultaneously too obscure (or uninformative) and too cluttering, and 2) citations should not provide detail that goes beyond allowing readers to locate the work cited. I have just demonstrated (18:58, 21 April) that the arbitrarily criterion of your first reason is absurd, and amounts to no more than an indirect form of WP:IDONTLIKEIT. And is not even entirely relevant here, as the current proposal does not require use of authorid. Regarding your second "reason": that is not a reason, that is only an unsupported personal opinion. What I have asked for, and what you have not given, is any authority or persuasive argument why anyone else should embrace your opinion. ~ J. Johnson (JJ) (talk) 20:10, 24 April 2016 (UTC)

Kanguole having nothing further to add, let us review Headbomb's objections. Particularly, that (19:36, 13 April)

1. it is "[not] Wikipedia's role to further clutter our editing window with ORCID data", and

2. "[t]he information is virtually pointless, and does no service to the reader."

He also elaborated a view that

3. any need for such information can be satisfied by following the doi back to "the journal's website."

Re #1: As I have previously stated (21:40, 13 April), if one's edit window has gotten too cluttered to readily follow the text then the biggest improvement is to pull out all of the citation templates. Not to dispense with them entirely, but to move them into their own section, where they can be organized neatly and without clutter, then use short cites in the text. ORCID data is not more cluttering than any other data used in a citation, and cutting that out on the basis of "clutter" is as sensible as cutting off one's feet because the blanket is too short.

Re #2: I and Andy have already explained how authoritative identification of authors, and the information thus available about an author, can be a valuable service to both the reader and the editor. At the very least this provides a basis for assessing an author's reliability. Such information generally includes a bibliography, which is extremely useful in seeing if the author has changed his position, or simply finding the latest work on the subject.

Re #3: Headbomb fails to recognize that not all works that might be cited are done under the auspices of a journal, and that not all journal websites provide author id links such as are proposed here. Having to follow such multi-step and possibly broken links disconnects readers and editors from potentially important information.

I believe all of the objections raised so far have been addressed, and are found lacking in substance. If there are no other objections I will proceed to a broader RfC. ~ J. Johnson (JJ) (talk) 20:47, 4 May 2016 (UTC)

Guess I'll start working on a longer version my oppose rationale. This should never be implemented. Wikipedia is not Wikidata. A once-per citation link to Wikidata via something like |wikidata=WIKIDATA-ID# is fine like we do with our other citation identifiers like |bibcode=1935JChPh...3..764W creates a link to the ADSABS database (Bibcode:1935JChPh...3..764W). Then you can shove all the ORCIDS for every author in there. Cluttering our articles with a bunch Smith, J^ORCID or Smith, J. or some such is not. Headbomb {talk / contribs / physics / books} 20:59, 4 May 2016 (UTC)

This looks like a strawman argument, in that the proposal is not to "clutter" the articles "with a bunch [of] Smith, J^ORCID". The specific use contemplated is more like: Smith, J.^o. Hardly a clutter. Nor, to anticipate Boghog's objection, a "sea of blue". ~ J. Johnson (JJ) (talk) 23:23, 4 May 2016 (UTC)

My objections above were never addressed. Boghog (talk) 21:03, 4 May 2016 (UTC)

Your objections were in a different discussion (#Sources (or authors) with a Wikidata item, but no article). Please state your objections to this proposal in this discussion. Keeping in mind that your feared "sea of blue" is really quite far fetched. ~ J. Johnson (JJ) (talk) 23:30, 4 May 2016 (UTC)

I too have commented on your proposal in the #Sources (or authors) with a Wikidata item, but no article thread. I support the goal you want to achieve, but I think that providing individual links for each and any of those ids isn't a good idea - at least not in citation templates. While inconvenient to use, I think that Wikidata is technically the better place to list all such ids, so that a single link per author to the corresponding Wikidata entry should be enough. That single link should either go directly to Wikidata or - better - to some special "intermediate page" (similar to Special:BookSources for ISBNs) which would function as smart frontend to the Wikidata entry. Details in the other thread. --Matthiaspaul (talk) 00:57, 6 May 2016 (UTC)

I agree entirely with Matthiaspaul; a single link to an intermediate page is the best approach. Peter coxhead (talk) 09:01, 6 May 2016 (UTC)

Instead of a Wikidata link per author, I would prefer Headbomb's suggestion of a single Wikidata link per citation. Added to the end with other identifiers, it wouldn't disrupt the citation details. Kanguole 09:43, 6 May 2016 (UTC)

I guess something like this will happen when citations are moved to Wikidata (in a couple of years?), similar to how we handle authority control info already.

However, I don't think it would be practical to try and create Wikidata nodes for citations manually now - moving data from citation parameters into the various entries of a Wikidata node is a task for bots and pulling data from such citation nodes back into a Wikipedia citation isn't yet supported by our citation templates as well. Without this infrastructure for display, edit and synchronisation purposes, this would almost instantly lead into chaos. Also, it would be extremely error-prone to manually set up links from citation objects to author objects.

However, we can already prepare for this by providing means to reliably distinguish between authors of the same name in citations. Basically, this is what Andy's and JJ's proposals are about.

Since a link to a particular data object is also a unique id for identification purposes, a single link per author should be enough. If we have an article about an author, the link should go there. Through the authority control framework already implemented that article serves as a frontend to the underlying Wikidata entry, where the various external IDs are stored. However, we don't have articles for all authors (and never will have), but we could have Wikidata nodes for all authors. So, we need a single link per author.

Proposal a) |author-linkn= will link to either an article (perferred) or a wikidata node. The citation template could automatically detect the target site. If the target is Wikidata, our citation template could slightly change the display of the link so that users can distinguish them from normal article links without having to click on them. Or the citation template could use the link as a handle to link to a special intermediate page rather than Wikidata directly.

Proposal b) In addition to |author-linkn=, we could introduce a |author-wdidn= parameter holding a Wikidata node number. Since these parameters are mutually exclusive, we could simply ignore |author-wdidn= if |author-linkn= is present as well. Alternatively, we could use |author-wdidn= to display a single additional link following an author's name (perhaps in the superscript format suggested by JJ), so we'd have at most two links per author. We don't have to decide on the specifics now, because for as long as all this is handled inside of the citation template, the actual behaviour and display format could be easily changed alongside the long migration process to Wikidata. However, in order to avoid potential misuse, the |author-wdidn= parameter should neither accept free-flow text as argument (as suggested by JJ) nor a free-flow url, but only the raw Wikidata node number itself.

--Matthiaspaul (talk) 19:37, 6 May 2016 (UTC)

I think we all agree that holding citations in Wikidata is the ideal destination, but there are differing ideas on how to get there. One approach would be to build all the relationships between bibliographic items and uniquely identified authors and editors (and maybe later journals, books, publishers and locations) in citation templates on Wikipedia, and have bots transfer them to Wikidata at some point in the future. One drawback with that is that it would turn citations into a "sea of blue", but I'm more worried about the massive redundancy it would involve, and the associated maintenance cost. Some useful citations are used in dozens or even hundreds of articles, with a corresponding repetition of the citation templates. Of course there is a lot of redundancy now, with |author-link= and the like duplicated many times, but these are more transparent than Wikidata IDs, and thus more feasible for editors to create, check and maintain.

So I would prefer a migration path that started with defining a citation representation on the Wikidata end, followed by extending these citation templates to use it, and then providing tool support to make it easier for editors to create reusable citations (held on Wikidata). Kanguole 00:09, 7 May 2016 (UTC)

The idea that there should be only a single id link for each author that goes to an intermediate location (Wikidata or something) reminds me of an intense little discussion some years ago (I vaguely recollect it was to do with use of short cites) regarding single-click versus double-click, and that readers shouldn't have to be put to the trouble of that extra click. Well, this discussion, here and now, does not have to get into that. Anyone that does not like {{authorid}} is free to come up with something different. The proposal here is simply that some provision be made so that the citation templates can handle what ever means of linking to author information, regardless of whether it involves single or multiple links, single or multiple clicks.

The idea of an |author-wdidn= parameter (perhaps that would be better as |author-wdn=?) linking to Wikidata is interesting, and parallel to what I am proposing here, except for being more limited. Note (as I have pointed out elsewhere, 21:49, 18 April) that {{authorid}} could be easily extended to provide links to Wikidata, which removes any need for a separate parameter.

Headbomb has suggested an alternative of having a single link per citation for "all the ORCIDS for every author" (but let's not forget Scopus, ResearcherID, and the rest). Details of just how that would be done might be interesting, if he would so enlighten us. But why stop there? It would certainly remove a lot of clutter (and blue) from the citations - which has been the most prominent concern of some of you - if we had a single link to similarly cover all the source links, such as axriv, asin, bibcode, doi, eissn, isbn, issn, jstor, lccn, ocic, pmc, pmid, ssrn, and zbl. But until we have concurrence for such "single links", and perhaps some concept of how that would be done, provisioning of an |author-idN= parameter is the most reasonable way of accommodating things like ORICD links, however that might be done. ~ J. Johnson (JJ) (talk) 21:21, 8 May 2016 (UTC)

Wikidata should have all the possible information about everything in the cited work. Article identifiers (arxiv, bibcode, doi, jfm, zbl, ssrn, etc.) should remain on Wikipedia to facility WP:V, but the author and journal identifiers (coden, issn, ORCID, ResearcherID, etc. should be on Wikidata). Each citation can then have a |wididata=WIKIDATA-ID# which would produce a link at the end of the citation

Smith, J. (2010). "Article of Stuff". Journal of Stuff 50 (4): 40 doi:10.1234/56789. SSRN 0123456798. Wikidata:Q123456798.

Headbomb {talk / contribs / physics / books} 21:36, 8 May 2016 (UTC)

Golly, gee whiz, and you were complaining about clutter?? Add an author-link, and a url for the article, and it is just about solid blue. But never mind that. Please show me: where a citation has multiple authors, how is the reader to find the information for a specific author? Can you show us how that linkage is to work? And what it will take to set that up? ~ J. Johnson (JJ) (talk) 22:23, 8 May 2016 (UTC)

References are for supporting the fact claimed, and the information is there to help people find works (book, journal articles, etc.), not for storing information about authors. If you want to find information about authors, you can look up the article in various databases (or go to Wikidata). You might as well ask "But how is the reader to find at what institution a specific author produced his work" or "What about the author's contact information?". This isn't the role of citations. Headbomb {talk / contribs / physics / books} 23:25, 8 May 2016 (UTC)

@Headbomb:, if I understand correctly you propose to migrate authorlinks for a reference to a Wikidata entry for that reference, leaving only a single link to Wikidata from the Wikipedia article that has that reference. I think this is a bad idea. In particular, it breaks the "what links here" feature (when used from the author's Wikipedia article) for finding Wikipedia articles that cite or otherwise mention that author, something I find very useful. It would also orphan a lot of articles. Or do you not include authorlinks as an "author identifier"? —David Eppstein (talk) 22:38, 8 May 2016 (UTC)

Author links aren't author ids, just like journal links aren't ISSNs, or that a link to Alpher–Bethe–Gamow paper isn't a doi. I'm not talking about changing the behaviour of what is or isn't wikilinkable. I'm talking about author IDs similar to ORCID and the like (or individual author links to Wikidata entries). Headbomb {talk / contribs / physics / books} 23:13, 8 May 2016 (UTC)

I agree that author links (wikilinks to articles) are not "author ids". And that wikilinking of author's names via |author-link= is behavior we should not be altering. Though I would mention that if the author-link processing was smart to exclude any kind of templating from the wikilink, but append it instead, then what I have proposed here would not be needed. But probably better to have separate parameters lest there be any confusion.

On your other point, you still fail to appreciate that citations are not simply identifying where something is stated, but very much a matter of who is responsible for a statement. That is the essence of reliable sources. In a given instance an editor's assessment of the reliability and credibility of a statement is based largely on the reputation and other characteristics of the author. Seeing (for instance) a list of the author's work helps to establish (or not) that author's area of expertise. (And as I have previously said, is very helpful in finding other works on a topic, and especially any recent items that might not be easy to find.)

Note that I am not proposing that "information about authors" should be "stored" in a citation, only that we should permit links to such information. Now you have suggested a single link per citation for all authors, but you have yet show how this would work for multiple authors. And I point out that your example (above) uses twenty characters for a single author, which is considerably more clutter and "blue" than having two or three characters (and small ones at that) for each author, as I have demonstrated. ~ J. Johnson (JJ) (talk) 20:58, 9 May 2016 (UTC)

I can only say that I agree with others here that citations are used to specify where the source is located, not to link to biographies or lists of works, and I am not persuaded that adding individual author ids is the way forward. (I find the edit summary "Citations are pretty much worthless if we don't know the authors" absurd, I regret to say.) Peter coxhead (talk) 21:24, 9 May 2016 (UTC)

Then perhaps that needs further discussion. To demonstrate the point, consider this: what if there was no author attribution. E.g., what if in some field authorship (and editorship) was simply ignored, and all contributions were anonymous. How would you determine whether some work was reliable? ~ J. Johnson (JJ) (talk) 23:34, 9 May 2016 (UTC)

(EC) What do you mean I haven't shown how this would work for multiple authors? Everything would be stored in Wikidata. All authors, all full names (even if we cite them as Smith, J. on Wikipedia), author affiliations, emails, author ids, etc. You get one Wikidata entry per citation, containing all this information. As for citation being worthless without authors, that's patently not true, given several journals omit authors entirely. If I give you Physical Review Letters 116, 191301 (2016), you can find the citation. If I give you doi:10.1103/PhysRevLett.116.191301 you can find it. If I give you Adam R. Brown et al, you can't find it, even if you knew which of the many Adam R. Brown in the world I was referring to. Headbomb {talk / contribs / physics / books} 00:24, 10 May 2016 (UTC)

Your example (above) is for "Smith, J." - a single author. Now perhaps you mean that your dummy WD link would somehow expand out for multiple authors, but so far you haven't shown how that would work (not even dummied); you're just waving your hands around.

You have misunderstood what I mean by no author attribution. You have given a short citation that does not include the authors in the citation, but that work does have identified authors (Adam R. Brown, Daniel A. Roberts, Leonard Susskind, Brian Swingle, and Ying Zhao) to which the work is attributed, who take responsibility for the work, and on whose repuations the credibility of the work ultimately rests. What I have proposed as a thought-experiment is a situtation where all work is anonymous, where we do not know, nor can know, whoe work it is. How would you determine whether such work is reliable when you don't know who the authors are? ~ J. Johnson (JJ) (talk) 23:14, 10 May 2016 (UTC)

Are you legitimately unable to understand the phrase Everything would be stored in Wikidata. All authors, all full names (even if we cite them as Smith, J. on Wikipedia), author affiliations, emails, author ids, etc.? Or unable to fathom that citations with multiple authors would work just the same, except would have more than one author listed? Like Smith, J., Thomson, K. (2010). "Article of Stuff". Journal of Stuff 50 (4): 40 doi:10.1234/56789. SSRN 0123456798. Wikidata:Q123456798? And that the Wikidata entry would contain the full names of both authors?

As for as your anonymous example, you look at the publication or its publisher. Anonymous paper hosted in a random forum? Likely unreliable. In The Economist (in which all articles are anonymous, last I heard)? Likely reliable. Headbomb {talk / contribs / physics / books} 23:29, 10 May 2016 (UTC)

So every statement and assertion in The Economist is absolutely true? And to the extent that any journal vouches for reliability of what they publish, would they do so if the did not who the authors are? What if someone responds that the original assertion is in error – do you have to wait for The Economist to sort out who is right? Would it make any difference to you if you knew who was making the several statements? Perhaps you should take a little more time to think this out. ~ J. Johnson (JJ) (talk) 23:52, 10 May 2016 (UTC)

Ok, at this point you're clearly not interested in having a dialogue. Go troll elsewhere. Headbomb {talk / contribs / physics / books} 00:02, 11 May 2016 (UTC)

Bullshit. I am trying (and I think very patiently) to address every objection, concern, and misunderstanding you (and others) may have so that we might resolve these matters. If you can no longer support your views you should consider whether they ought to be modified. ~ J. Johnson (JJ) (talk) 21:17, 11 May 2016 (UTC)

My opposition is not due to a misunderstandings, but rooted in the fact that author ids DO NOT BELONG IN CITATIONS AND ARE COMPLETELY POINTLESS INFORMATION THAT SERVES NO PURPOSE WHATSOEVER IN HELPING READERS TO VERIFY THE INFORMATION. You, on the other hand, can't understand basic sentences, apparently. That information belongs in Wikidata, not in our reference sections. Headbomb {talk / contribs / physics / books} 02:04, 12 May 2016 (UTC)

Your example of a citation that omits specifying the authors is a misunderstanding of my hypothetical of totally anonymous works.

If shouting makes you feel better, fine, but it's not persuasive, and does not advance any dialogue. ~ J. Johnson (JJ) (talk) 21:49, 12 May 2016 (UTC)

@J. Johnson: to return to the question you asked me. If I look at UCERF3, I have no idea who any of the authors of the sources given are, and I don't care. I do care that the sources are papers in reputable scientific journals; they might just as well be anonymous. I work regularly with secondary sources such as the World Spider Catalog or the World Checklist of Selected Plant Families, where authorship is irrelevant. Peter coxhead (talk) 21:49, 11 May 2016 (UTC)

You are less concerned with authorship because you are relying on the journal editors at both the primary (original publication) and secondary levels, who have already passed on the authors' qualifications, etc. But make no mistake about this: the authority of all scholarship, scientific study, and persuasive speech of any kind ultimately rests on the authority or credibility of the original author(s). If an editor receives (say) a paper with much arcane formulae and difficult to follow argumentation that asserts a modification to some physical constant, it makes a big difference whether the author is some reclusive but brilliant mathematician, a thousand-member team at CERN, or the self-styled King of Bosnia. If the paper is entirely anonymous - that is, the editor knows absolutely nothing about the author - then it will most likely, and most rightly, not be published.

Authorship is never irrelevant, even you if don't have to deal with it directly. But what you refer to is where something has become so well established and generally accepted that acceptance is no longer contingent on the original author. Where WP editors cite recent sources, the claims are not yet well established, so it is often helpful to know more about the authors. More journals are coming around to providing ORCID links as a service for their readers. It would be service for our readers (and editors) to be able to do the same. ~ J. Johnson (JJ) (talk) 21:56, 12 May 2016 (UTC)

Authorship is always irrelevant, except in the cases where it serves to dismiss the writings outright as the ramblings of quacks (and those won't have ORCIDs). I have never ONCE relied on the name on top of a paper to tell me if it was a good one or not. The quality of a paper, or of a book, or a blog piece depends on the quality of the arguments made, not the person who makes them, e.g. Kary Mullis. That he has a Nobel Prize utterly fails to impress me to the point that I'll think AIDS is fake, as is climate change, or that astrology is legit. ORCID serves no purpose whatsoever to accomplish WP:V. Either way, the place of author ids isn't here, but on Wikidata.Headbomb {talk / contribs / physics / books} 22:08, 12 May 2016 (UTC)

Authorship is not only relevant, but necessary, to determine whether a self-published source passes the "established expert" exception to WP:SPS, part of Wikipedia's core policies. —David Eppstein (talk) 22:20, 12 May 2016 (UTC)

That might be the only exception, but the author's name and the source itself will be plenty sufficient to establish if this is an expert or not. No need for an author ID in that case either. Headbomb {talk / contribs / physics / books} 22:46, 12 May 2016 (UTC)

But you will accept that as an exception? Good. Now consider that author's names often come in different forms, and not every article or book is listed by Google Scholar. Consider also that experts often modify their opinions, which you might have difficulty finding without access to a fairly complete bibliography. ~ J. Johnson (JJ) (talk) 23:43, 12 May 2016 (UTC)

We simply shouldn't normally be doing this kind of investigation into authors' credentials. It's completely against the spirit of WP:NOR. We should be relying on secondary sources to establish notability and importance, and then using original sources only to fill in details. Any article whose sources entirely rely on WP:SPS fails to meet the required criteria for inclusion in Wikipedia; it should never be the key source.

In any case, however interesting to some of us, this part of the debate is irrelevant. The choice is between (1) adding an id for every author (2) adding an id for the citation as a whole which then links to the authors. I've seen no convincing argument as to why (1) is best. There are lots of disadvantages to (1), including clutter, avoiding repeated links – it would be tricky to make sure that an author is only linked once, as well as the inappropriate suggestion that Wikipedia evaluates the importance of authors. Peter coxhead (talk) 08:22, 13 May 2016 (UTC)

An argument that has been presented for (1) is that Wikidata is not ready, so we should create the links in Wikipedia. This seems very weak: Wikidata is surely much better suited to this sort of linkage. Kanguole 08:51, 13 May 2016 (UTC)

I agree; there's no reason to mess up citation formatting here just because Wikidata isn't ready yet. It's not a vital or urgent issue. Let's wait until it can be handled there. Peter coxhead (talk) 09:36, 13 May 2016 (UTC)

I agree that normally we don't investigate author's credentials. But by the same argument we should dispense with citations because normally our readers don't investigate the sources. Cases do arise (such as determining reliability) where an author's credentials are relevant, and making such an assessment is no different than assessing the relevance or notability of any source material. That is not prohibited by WP:NOR; it is part of the ordinary discretion expected of editors. Note that I do not maintain that an author ids must be added in all cases, only that there be some means of adding an id when one is needed.

Your description of alternative (1) is incorrect: this is not about "adding an id for every author", it is about adding links to authors (may be only one) individually. As to "clutter" (as I have previously noted): having a single, superscripted small-letter appended after half-a-dozen author names is a third less clutter (and less blue) than the twenty characters of Headbomb's "Wikidata:Q123456798" example. Your objection of "clutter" really should run against the provided Wikidata example.

As for "avoiding repeated links": what are you talking about? ~ J. Johnson (JJ) (talk) 21:36, 13 May 2016 (UTC)

It's the same problem there is with |author-link=. A given author should only have one id link in the article (more definitely is clutter). But this is difficult to ensure when refs get added incrementally and change position as the text changes.

We're not persuading one another. No more on this subject from me. Peter coxhead (talk) 21:56, 13 May 2016 (UTC)

Multifaceted issues take more time and work, but I am willing to stay the course.

The sub-issue you raise here seems to be a variant of the MOS:OVERLINK issue, where an article is deemed to contain an "excessive" number of links. In regards of |author-links=, and also in wikilinking publishers and cities of publication, there is a broader issue: where a wiki-notable author (or publisher) appears more than once in a bibliography, it does seem excessive to have that name highlighted multiple times. E.g. (to strip matters to the essentials), something like:

• Std. "blue" links: John Smith ...; John Smith ...; John Smith ...; John Smith ...;

certainly seems excessive. But that is more of an issue with author-links. As an alternative (such as with {{authorid}}) consider:

• Superscripted: John Smith^o ...; John Smith^o ...; John Smith^o ...; John Smith^o ...;

Massively "less blue". So, yes, repeated links can be a problem, even amounting to overlinking, but that is a general problem, and more of problem with author-links (but not deemed a show-stopper!). It is actually less of a problem with the author id links shown here. And I can see other uses of an |authorid= parameter that could reduce the flood of blue created by author-links. In the issue raised here |authorid= is less of the problem, and more of an opportunity for a solution. ~ J. Johnson (JJ) (talk) 23:49, 14 May 2016 (UTC)

Oppose—at this time, I'm not in favor of adding author ID numbers to citations. Maybe in the future we can use Wikidata as a sort of intermediary repository for them, but we're far from that point. In the interim, adding these would add visual clutter without an immediate tangible benefit. (If there were such a benefit, other style guides would have determined a way to add such ID numbers to citations, but they have yet to do so.) ID numbers for the actual sources (ISBN, ISSN, OCLC, DOI, etc), yes, but not for authors. Imzadi 1979 → 07:03, 14 May 2016 (UTC)

"Other style guides" would be CMS, APA, MLA, etc., which is to say: the traditional guides to print media. And as print media lack a linking function, they would have to use the raw ID numbers, which definitely would "add visual clutter". But that is not what is proposed here. What I have proposed is a way to allow (among other possibilities) use simple, single-character hyper-links. ~ J. Johnson (JJ) (talk) 23:53, 14 May 2016 (UTC)

Strictly speaking, my statement is not quite accurate. The print media concept of a raised (super-scripted) asterisk (^*) is essentially a directive to self-mediate "linking" to a note, typically at the foot of the page. That is basically what the {{authorid}} template implements, a single-character superscript, except for having a different target. ~ J. Johnson (JJ) (talk) 22:40, 16 May 2016 (UTC)

I see that multiple authors in the author field is classed now as an error. However this is my preferred way of coding, as, in my opinion, attempts to break down authorship field more granularly are not the province of citation templates.

All the best: Rich Farmbrough, 17:19, 23 April 2016 (UTC).

I don't understand. You approve of citation templates listing different authors (granularity) but you disapprove of the formatting of such granularity. (a) Why? (b) isn't this contradictory? 65.88.88.69 (talk) 18:01, 23 April 2016 (UTC)

One of the primary purposes of citation is to identify the authors of a source. And not as as mass of confused text, but clearly and even individually. How is this not in "the province of citation templates"? ~ J. Johnson (JJ) (talk) 22:32, 23 April 2016 (UTC)

Lumping all the authors together is contrary to the cite template documentation. Jason Quinn (talk) 07:26, 24 April 2016 (UTC)

Piling on, |ref=harv doesn't work and can't work unless the authors are separated. (Attempts to do the separation automatically in software are doomed because of the too-great variety of names and of ways that template users might try to format them.) —David Eppstein (talk) 07:32, 24 April 2016 (UTC)

And let us not forget COinS, or any other function that depends on correct metadata. Though I will say one thing in favor of copy-pasting entire author lists (often including footnote numbers or symbols) into |authors=: it is an up-front indication that the rest of the article may suffer from the same lack of care and diligence. ~ J. Johnson (JJ) (talk) 20:20, 24 April 2016 (UTC)

I entirely agree that multiple authors should be split up, for all the reasons noted above. However, it seems that the many of those participating in this discussion would regard such splitting up as contrary to WP:CITEVAR, which they regard as preventing changes in coding without first obtaining consensus. Peter coxhead (talk) 22:06, 24 April 2016 (UTC)

Not relevant. The OP is complaining because he is using a single-person parameter (|author=) where he should clearly have used the multiple-person version (|authors=). And then makes the contradictory statement that "granularity" regarding authors is not the province of CS1. 65.88.88.200 (talk) 13:31, 25 April 2016 (UTC)

Because of its free-form nature, we should be discouraging the use of |authors=. The content of |authors= is not made part of the citation's metadata because we can't necessarily parse the parameter's contents into the appropriate metadata keywords rft.au, rft.aufirst, rft.aulast.

—Trappist the monk (talk) 14:05, 25 April 2016 (UTC)

For sure. Did we not deprecate |authors= along with |coauthors=? ~ J. Johnson (JJ) (talk) 21:34, 25 April 2016 (UTC)

Not deprecated, nor imo should it be. Neither should |editors= or |people=. Although I've rarely used them, not everything can be put in a straightjacket. 72.43.99.146 (talk) 23:48, 25 April 2016 (UTC)

"Straightjacket" is hardly a neutral, dispassionate characterization; I reckon that just slipped out. As to any difficulties, problems, or special cases for which stuffing multiple authors into an |authors= parameter is a reasonable solution: could you offer any examples? ~ J. Johnson (JJ) (talk) 00:30, 27 April 2016 (UTC)

It's spelled "straitjacket". --Redrose64 (talk) 09:54, 27 April 2016 (UTC)

I'm glad that the meaning was understood despite the spelling. But I disagree that "straitjacket" is not apt. Because deprecating these parameters is forcing a narrower set of useful options. Pointing this out is neutral. If there is a problem with the metadata, then (a) change the coding, not editor's behavior (b) deprecate the metadata, not a parameter that helps editors present sources. These parameters can be useful, with authors that are known by different names or by pen-names. Such names can be abandoned, then the same work can be republished listing the same author with a diferrent name. This could impact verifiability, as at first glance, the author (of the earlier edition) doesn't exist. And there is no |orig-name=. Putting all such names into |author= or |authorn= is semantically incorrect – there are many names, not many authors. |authors= is not correct either, but it is better (author-separation is not explicit). Also the parameters can be useful when roles may have to be designated, for example with authors of creative works such as (director), (producer), (scriptwriter) etc. Using |first= to add these roles is also semantically incorrect. If and when an |author-id= is established, then these free-form parameters can be revisited – assuming that |author-link= or |author-id= are filled-in, and correctly so. 65.88.88.126 (talk) 14:18, 27 April 2016 (UTC)

The history of the jacket is certainly in the sense of "confining to a narrow space". But I tend to be ambiguous, even tolerant, on whether it might equally well mean "to make straight".

It is not clear just what you mean. (Specific examples might help.) In cases where an author is known by various names, there are various ways of handling that. (And perhaps with author-ids, but that's a different discussion.) But question is not about putting all such names into |author=, but of putting multiple authors into |authors=. So like I asked before: do we have any examples of where stuffing multiple authors into an |authors= parameter is a reasonable solution? ~ J. Johnson (JJ) (talk) 21:09, 27 April 2016 (UTC)

A single option for anything, is a situation 100% made straight. But some situations are not reducible to that. There can conceivably be situations where a source is cited with a no-longer used author name. If one tries to verify that source, a different author name will pop-up and this may lead to unnecessary discussion in the article's talk page, and then here. Especially if there is no way of immediately getting the particulars of an author through following |author-link= or |author-id=. The temporary solution for me would be to use |authors= for the reasons stated above, to hold both author names. And as also stated above, there is the other situation, where author roles may be indicated. 72.43.99.146 (talk) 00:30, 28 April 2016 (UTC)

Please read more attentively, as you seem to be missing the essential difference between authors, and names of authors. E.g., when you say that |authors= can "hold both author names", it is evident you are talking about multiple names for a single author. But!! 1) All of that, being applicable to a single author, could be done in the singular |author1=. (Or even better in |first1= and |last1=.) 2) what we are discussing here is not a single author with multiple names, but the use of the plural |authors= parameter for multiple authors. E.g.: |authors=Smith, M.; Jones, Jim, Jr.; Miller, Richard. Do you understand that? ~ J. Johnson (JJ) (talk) 19:18, 28 April 2016 (UTC)

Adding more than one anything in a single-attribute parameter is both semantically and programmatically incorrect. Conflating 2 names/identities in that field falls under this error. Using "authorn" is also wrong, as the explicit author separation is similarly incorrect. But |authors= is free-form. By definition more flexible, and though not perfect, can handle such situations better. Here's some pseudocode: {{cite book|authors=Penname [pen-name of Author]|title=Title}}. 64.134.69.85 (talk) 22:28, 28 April 2016 (UTC)

Your example, of the attributes of a single author, illustrates some confusion. Please note: the "s" in |authors= is the plural form, not the possessive form. (|author's=??)

I have no idea what you mean by "the explicit author separation" you infer to "authorn". Perhaps another example would help. ~ J. Johnson (JJ) (talk) 21:07, 29 April 2016 (UTC)

The term "authors" despite the plural form may include only a single author (any author over 0). "Author" indicates a singular value: just 1 author. When singular authors are used, such as "author1", "author2", etc. the emphasis is on their uniqueness. However "authors" can be used to indicate commonality. We are talking about the best option in a batch of bad ones. 184.75.21.30 (talk) 23:36, 29 April 2016 (UTC)

Sorry, no, you are talking nonsense. "Authors" is plural, and neither the term nor the so-named parameter suggest any element of commonality. Indeed, neither does your example (did you perhaps mis-take "commonality" for "property"?), and I suspect your grasp of the concept is weak. Your comments have demonstrated nothing but your own non-understanding. As that seems innate it would be a waste of time to continue this discussion. ~ J. Johnson (JJ) (talk) 20:30, 30 April 2016 (UTC)

Not so crazy. Lumping all the authors together is required by |vauthors= ;-) Boghog (talk) 20:58, 30 April 2016 (UTC)

The |vauthors= parameter does not require "[l]umping all the authors together", it is the means by which that is achieved in accordance with the Vancouver style of citation, and which is done on the expectaton of a specific structuring that is not expected with |authors=. Aside from that, "vauthors" is entirely parallel with "authors" in that the "s" means the plural form, not the possessive form. Nor whatever fantastical form this anonymous IP has in mind.

And now that has been clarified perhpas someone would hat this entire digression, from the first "straightjacket" comment of 23:48, 25 April.

Whoops! That is my unsigned edit of 00:47, 2 May 2016‎ ~ J. Johnson (JJ) (talk)

If people wish to see multiple authors listed in an "authors" parameter, then I can understand that. I think, though, that cite templates should be easy to use. It is really cool that people can rip data from WP in structured ways, but what I primarily care about is making the encyclopaedia easy to read and write. All the best: Rich Farmbrough, 00:07, 5 May 2016 (UTC).

CS1 templates are not mandatory. Another thing that is not mandatory is that they should be easy to use. It would be nice if they were, but there is no design mandate that they absolutely must be. The main defining element of style is consistency, not ease of use. And in any case, all that is irrelevant, because that is not the matter you raised in your OP. 72.43.99.146 (talk) 00:19, 6 May 2016 (UTC)

I suspect many editors want an "authors" parameter because they don't want to go to the trouble (as in "it's too hard!!!!") to parse out individual authors. Which gets into an interesting consideration: anything really easy to do would already be done, just by the natural course of things. (Like gravity concentrating hydrogen to make stars happen.) But any thing desirable that is not yet done will require some effort to achieve. If "easy to write" was our primary criterion then all kinds of pap would result, of little interest or worth to read. As encyclopediasts our primary goal is to inform. As a secondary criterion we hope this won't be any harder than necessary. ~ J. Johnson (JJ) (talk) 20:33, 5 May 2016 (UTC)

I would not object if we'd deprecate |authors= for as long as we don't delete it. It is convenient to have in the first run of a conversion job of free-flow citations to use the citation templates (or when translating citations from other Wikipedias), in particular in articles outside one's own interest or when in a hurry. Deprecating the parameter would not be in conflict with this usage, we could even track its usage in some maintenance category. Perhaps someone will write a tool to semi-automate the process of splitting |authors= into separate parameters. --Matthiaspaul (talk) 00:35, 6 May 2016 (UTC)

No, "authors" should not be deprecated, for reasons given above. This goes beyond presentation, which is the province of style (and of this page), and into the substance of citing sources. Any discussion of deprecation should happen there, not here. Style-wise, "authors" can present the citation exactly as another person-related parameter, assuming the doc gives proper guidance. The real problem here is that the doc is not up to par. 207.237.37.100 (talk) 15:58, 6 May 2016 (UTC)

If |authors= presents the citation in exactly the same way as splitting up the information into |last1=, |first1=, etc., where's the problem in deprecating |authors=? Anyone can use this short form and then other editors can tidy up behind them. Peter coxhead (talk) 16:38, 6 May 2016 (UTC)

Fine, let's deprecate the individual params then. But let's get serious. There's place for both. 160.79.53.242 (talk) 22:42, 6 May 2016 (UTC)

With careful hand-crafting one can use |authors= in way that appears the same as using |last/first=. But why go to that much trouble instead of just using last/first? Not to mention that CITEREFs, COinS, and any other current and futuristic uses of the metadata are broken. In the long run last/first is indeed easier than using |authors=, which is little more than a means of shifting some necessary work on to someone else. I see no place for it. ~ J. Johnson (JJ) (talk) 23:24, 6 May 2016 (UTC)

This is useless discussion. If anyone feels that "authors" should be deprecated, start a proper discussion with proper reasoning at Wikipedia talk:Citing sources. Since the presence or absence of any person parameter can be presentation-neutral, this is not a matter of presentation, and therefore not a question of style. Which is what CS1 is all about. If and when "authors" is consensus-decided that has no useful function, then the implementation of the new consensus will be discussed. That discussion belongs here. 72.43.99.146 (talk) 13:15, 7 May 2016 (UTC)

WP:Citing sources is an overview of citation, is neutral regarding whether one uses cs1 style or cs2 style templates, or, for that matter, any templates at all, and says nothing about the use of template parameters. On the other hand, this page details how to use the cs1 templates generally, covering specific parameters in detail. And while this topic is termed "Citation Style 1", the use of these templates and their parameters goes beyond mere presentation, it also goes to the integrity of the information contained. If you can show us a better place to discuss how to use these parameters, please do. And of course, if you find this discussion to be totally useless, you are free to ignore it. ~ J. Johnson (JJ) (talk) 22:44, 7 May 2016 (UTC)

Where was it agreed that all instances of |author= should be replaced with |authors=? Dcirovic (talk · contribs) is making edits like this at the rate of 12 per minute, or one every five seconds, and are effectively acting as a WP:BOT despite not being a bot account. I left a talk page message, but they continue unabated. --Redrose64 (talk) 08:20, 14 May 2016 (UTC)

There has been no such agreement and that kind of fix is wrong because it fixes nothing and, because the content of |authors= is not made part of a cs1|2 template's metadata, single name |author= which were correct and were creating author metadata are no longer doing so.

For whatever it's worth, I've added text to Category:CS1 maint: Multiple names: authors list so that, on the off chance that other editors read it, this kind of brute-force change might in future be avoided.

—Trappist the monk (talk) 09:02, 14 May 2016 (UTC)

Please see Category talk:CS1 maint: Multiple names: authors list. I reverted Trappist's copyedit for reasons given in that page. 65.88.88.126 (talk) 16:23, 16 May 2016 (UTC)

As a possible summary: are we generally in accord that (aside from the structured vauthors parameter), it is preferable to not put multiple authors into a single |authors= parameter? ~ J. Johnson (JJ) (talk) 23:40, 17 May 2016 (UTC)

Cite magazine comparison

Wikitext	{{cite magazine|date=1986|first=Harald E.L.|last=Prins|number=1|pages=4-9|postscript=(magazine of the Fort Western Museum).|title=The Most Convenientest Place for Trade; a Discussion of the Kenibec/Cushnoc Controversy|volume=3|work=The Kennebec Proprietor}}
Live	Prins, Harald E.L. (1986). "The Most Convenientest Place for Trade; a Discussion of the Kenibec/Cushnoc Controversy". The Kennebec Proprietor. Vol. 3, no. 1. pp. 4–9(magazine of the Fort Western Museum).{{cite magazine}}: CS1 maint: postscript (link)
Sandbox	Prins, Harald E.L. (1986). "The Most Convenientest Place for Trade; a Discussion of the Kenibec/Cushnoc Controversy". The Kennebec Proprietor. Vol. 3, no. 1. pp. 4–9(magazine of the Fort Western Museum).{{cite magazine}}: CS1 maint: postscript (link)

I found citation 5 in this revision of "Cushnoc Archeological Site" misuses |postscript=. Should there be a test for the use of a punctuation character? I would suggest that the module output 1) an error if this parameter is set to anything more than one character in length, and 2) a maintenance notification if the character is not a punctuation mark (in English). --Izno (talk) 17:13, 16 May 2016 (UTC)

For previous discussions (with no change in the documentation or the template's behavior), see this previous discussion from December 2014 and the end of this discussion from July–October 2015. – Jonesey95 (talk) 21:00, 16 May 2016 (UTC)

Just counting on my hands, it looks like there is consensus for this idea, but there may need to be work done post-implementation into the module (StoryOfWikipedia). --Izno (talk) 22:47, 16 May 2016 (UTC)

Arguably, this and similar problems may be avoided by using correct nomenclature. In its current iteration it is obvious that |postscript= does not stand for postscript, but for terminal punctuation. Renaming the parameter according to its purpose would be the first (and perhaps the only necessary) step. 72.43.99.146 (talk) 14:10, 17 May 2016 (UTC)

(edit conflict) While I have not used |postscript= for this purpose myself (although I was tempted to do so many times), a lot of editors use it to store visible notes related to the citation. I would therefore oppose enforcing stricter checks for the |postscript= parameter for as long as we don't have a |comment= or |note= parameter as was repeatedly proposed by various editors already and is supported by the citation templates in some other Wikipedias. The contents of that optional parameter would be shown following the optional output of the |quote= (and the proposed |script-quote= and |trans-quote=) parameter(s), and ideally it would be framed in parenthesis (like: "(<text>)" or "(Note: <text>)" or "(NB. <text>)"), so that the |postscript= parameter can still be used to define the leadout character in order to blend it in with the text surrounding the template. Keeping the various texts in separate parameters (and spans) would allow us to bidir-protect the contents of the script parameter, possibly add language metadata, dynamically filter specific contents based on output media (monitor, screen, printer, data stream), style (CSS) and user preferences, and let the template adjust the leadout semi-automatically. --Matthiaspaul (talk) 14:41, 17 May 2016 (UTC)

Please provide a showing of community consensus to support |comment= or equivalent. I have seen a number of users advocate such a parameter but no WP:RFC to support it.

Regardless, this is a misuse of the parameter. Any further discussion regarding |comment= deserves its own thread (again); blocking the implementation of a test for misuse of this parameter for a "I want this other parameter" would be pointed though not-particularly disruptive (and probably falls in the WP:OTHERSTUFF bucket of issues). --Izno (talk) 14:59, 17 May 2016 (UTC)

That was unduly confrontational. If a parameter is often abused as a workaround for missing functionality, it is reasonable to decide whether we want that feature, and if so to implement it so that the message can provide editors with a way forward, rather than just telling them they're wrong. Kanguole 16:02, 17 May 2016 (UTC)

I'm simply not going to reply, as I've already made the points I think I need to make to support a test for this parameter. The "new section" button awaits you, for your |comment= RFC. --Izno (talk) 16:23, 17 May 2016 (UTC)

Where a parameter is being misused it is likely useful to determine why. And if there is some need that is not being met then by all means some thing might be proposed. But a problem somewhere else does not justify misuse of this parameter. I support a restriction (and test) on the values supplied to |ps=. And I wonder if anything other than a comma or a period is valid. I think we should also consider whether the name should be changed. ~ J. Johnson (JJ) (talk) 23:32, 17 May 2016 (UTC)

Following up on the suggestion by 72.43.99.146: anyone have any ideas for a better name for this parameter, that would avoid the association with PostScript, and better communicate that this is about choice of punctuation? ~ J. Johnson (JJ) (talk) 19:29, 22 May 2016 (UTC)

The suggestion in Module:Citation/CS1/Suggestions for Italian parameter names are all wrong (except for "volume" and "titolo"). I have prepared an updated file with the name used in it.wiki, can I upload in the sandbox Module:Citation/CS1/Suggestions/sandbox ? --Moroboshi (talk) 07:13, 24 May 2016 (UTC)

Yep, that's what the sandbox is for.

—Trappist the monk (talk) 10:38, 24 May 2016 (UTC)

...and with a typical bad move I clicked the wrong link and updated the live suggestions module - I reverted my change and uploaded the change for Italian suggestions in the sandbox.--Moroboshi (talk) 12:39, 24 May 2016 (UTC)

Asking this here b/c I don't see it addressed on Category:CS1 errors: Vancouver style (so I was forced to guess).

The module prefers |vauthors=Gower Jr WR, Carr MC over |vauthors=Gower WR Jr, Carr MC (my emphasis) (see here and here). The latter emits an error and fails to render the r in Jr. However, I've seen that the latter is more common than the former (only anecdotally; I haven't done any research into their relative frequency). Are there any plans to allow this? If not, it should be stated somewhere on the error category.   ~ Tom.Reding (talk ⋅dgaf)  18:30, 17 May 2016 (UTC)

See here. cs1|2 doesn't understand generational suffixes. To render them properly, use the doubled parentheses to inhibit error checking for that name:

|vauthors=((Gower WR Jr)), Carr MC

—Trappist the monk (talk) 18:41, 17 May 2016 (UTC)

Thanks. I see that the purpose of the doubled parentheses is to allow for non-standard text. It's common to come across Jr & Sr, though, and it seems relatively straight-forward to unambiguously identify it in the author-string, i.e. it must exist either before or after first/middle initials ([JS]r\s)?[A-Z]+(\s[JS]r)?, which must exist after a last name [A-Z][a-z]+ to not emit an error. I can't speak to the difficulty/ease of implententation, however.   ~ Tom.Reding (talk ⋅dgaf)  19:09, 17 May 2016 (UTC)

Does Vancouver style even allow Sr./Jr./etc.? It seems strange to care only for the first and middle initials, but then want to know whether that is WR Jr. or WR Sr. I don't know what the rules for Vancouver are, but wouldn't it be just as well to just leave that stuff of? ~ J. Johnson (JJ) (talk) 23:12, 17 May 2016 (UTC)

Currently, |vauthors=Last Sr. FM is allowed (period is removed prior to rendering) but |vauthors=Last RM Sr isn't.   ~ Tom.Reding (talk ⋅dgaf)  13:33, 25 May 2016 (UTC)

I just noticed #generational suffixes and Vancouver style. Thanks!   ~ Tom.Reding (talk ⋅dgaf)  13:39, 25 May 2016 (UTC)

|vauthors=Last Sr. FM is malformed. Vancouver style requires that the suffix to be placed at the end after the initials. Module:Citation/CS1 treats 'Sr.' as part of the last name. The disappearance of the dot in rendering is part of the reason I want to rewrite the name handling code. At the next update, a dot appearing in any |vauthors= parameter will cause an error message:

{{cite book/new |title=Title |vauthors=Last Sr. FM}}

Last Sr FM. Title. {{cite book}}: Vancouver style error: punctuation in name 1 (help)

—Trappist the monk (talk) 15:37, 25 May 2016 (UTC)

One of the obstacles to adopting |vauthors= usage on pages using Vancouver style authors seems to be no option for a trailing period at the end of the rendered author list (even though a period appears after the date). I'm not sure why this extra period would be desired (pinging Sasata, who supports it), but is it something that we want to implement?   ~ Tom.Reding (talk ⋅dgaf)  19:05, 24 May 2016 (UTC)

I'm not sure what it is that you are asking. Puntuation in the rendered citation is the responsibility of the cs1|2 templates. {{cite book}} and the other cs1 templates will terminate the author name list with a period regardless of the punctuation that separates the individual names. Likewise, {{citation}} (or cs1 templates with |mode=cs2) will terminate the name list with a comma:

cs1:

{{cite book |title=Title |vauthors=Last FM, Last FM}}

Last FM, Last FM. Title.

cs2:

{{citation |title=Title |vauthors=Last FM, Last FM}}

Last FM, Last FM, Title

|mode=cs2:

{{cite book |title=Title |vauthors=Last FM, Last FM |mode=cs2}}

Last FM, Last FM, Title

—Trappist the monk (talk) 19:22, 24 May 2016 (UTC)

This is what is desired (though not by me):

Mushrooms of North America in Color. 1995. {{cite book}}: Unknown parameter |authors= ignored (help)

Which unfortunately has to use |authors= to get the job done, since |vauthors= emits an error when a trailing period is present. If this were doable via |vauthors= (or some other parameter (|postscript= comes to mind)) that would solve the problem (removing the desire to have a trailing author period when |date=/|year= is listed would solve it too, but that doesn't seem possible given the adamance of the above linked discussion).   ~ Tom.Reding (talk ⋅dgaf)  19:38, 24 May 2016 (UTC)

Examples:

{{cite book |title=Title |vauthors=Last FM, Last FM|year=1999}}

Last FM, Last FM (1999). Title.

{{cite book |title=Title |last=Last|first= FM|last2= Last|first2= FM|year=1999}}

Last, FM; Last, FM (1999). Title.

{{cite book |title=Title |authors=Last FM, Last FM|year=1999}}

Title. 1999. {{cite book}}: Unknown parameter |authors= ignored (help)

|authors=, |last1= etc., and |vauthors= are handled the same. Adding a period to the list of authors will make it inconsistent with the CS1 style, such as it is. An editor who wants a period after the author list and before the year may need to use hand-rendered citations.– Jonesey95 (talk) 22:10, 24 May 2016 (UTC)

(edit conflict) great minds?

An issue raised in the linked discussion appears to be consistency. So, checking the four flavors of author parameter to make sure that they are consistent with each other; first without a date:

Last FM, Last FM. vauthors.

Last FM, Last FM. "author". Work. – {{cite news}} just as a check

authors. {{cite book}}: Unknown parameter |authors= ignored (help)

Last, FM; Last, FM. last/first.

and with date:

Last FM, Last FM (1995). vauthors.

Last FM, Last FM (1995). "author". Work. – again {{cite news}}

authors. 1995. {{cite book}}: Unknown parameter |authors= ignored (help)

Last, FM; Last, FM (1995). last/first.

it appears to me that the templates are consistent in the application of terminal punctuation

The same using {{citation}}; first without date:

Last FM, Last FM, vauthors

Last FM, Last FM, author

authors {{citation}}: Unknown parameter |authors= ignored (help)

Last, FM; Last, FM, last/first

and with date:

Last FM, Last FM (1995), vauthors

Last FM, Last FM (1995), author

authors, 1995 {{citation}}: Unknown parameter |authors= ignored (help)

Last, FM; Last, FM (1995), last/first

again, these are consistent with each other. But, this test does reveal a CITEREF anchor bug, so this discussion isn't for naught.

If consistency is the issue, that suggests that an article is using more than one kind of citing mechanism. If cs1 is chosen, then all references must use the cs1 templates. It is the same for cs2. If the desired formatting is not directly supported by cs1|2, then use an appropriate alternate template or free-form citations are in order.

cs1|2 have their roots in a variety of published style guides, one of which is APA style from which they get their parenthetical dates following after the author list. At some point in the development of cs1|2, a decision was taken to omit author list terminal punctuation when the template has a date. Without significant discussion that reaches a determination to change all author, editor, translator, and contributor list parameter processing, it seems improper to change only |vauthors=.

—Trappist the monk (talk) 22:30, 24 May 2016 (UTC)

Excuse my ignorance, but are hand-rendered citations that Jonesey95 mentions, and free-form citations that you mention referring to the same thing, and to, for example, something like <ref>Last FM, Last FM. (1999). Title.</ref>?   ~ Tom.Reding (talk ⋅dgaf)  13:58, 25 May 2016 (UTC)

Yes, that is what is meant: citations that do not use templates. – Jonesey95 (talk) 14:43, 25 May 2016 (UTC)

Yep.

—Trappist the monk (talk) 15:38, 25 May 2016 (UTC)

It seems that DOIs are rendered as:

 https://dx.doi.org/10.1029%2F2009JD012104
 https://dx.doi.org/10.1093%2Fref%3Aodnb%2F31543

These links are from Lowest temperature recorded on Earth, using {{cite journal}}, and Nikolaus Pevsner, using {{cite ODNB}}. In both, a slash (/) is rendered as %2F, and in the second a colon (:) is rendered as %3A. Both use the unencoded form (10.1029/2009JD012104 & 10.1093%2Fref%3Aodnb%2F3154) in the wikitext. It seems, as far as I can tell, to affect anything in the CS1 family.

Is there a particular reason for doing this? URLs, etc, that come through CS1 seem to be unencoded. Andrew Gray (talk) 11:52, 29 May 2016 (UTC)

Because a DOI can use pretty-much any character. The same is not true for urls so those characters not allowed in a url must be percent encoded.

—Trappist the monk (talk) 12:22, 29 May 2016 (UTC)

Aha, thanks - should have realised this, I had horrible problems writing a script which interpreted DOIs some months back :-). There's an unexpected fringe benefit of this - potentially makes it easy to identify DOIs used directly rather than called through a template, without having to parse the wikitext. Andrew Gray (talk) 12:29, 29 May 2016 (UTC)

Module:Citation/CS1 is not supposed to create a CITEREF anchor id from |authors=:

{{citation |title=authors |authors=Last FM, Last FM}}

'"`UNIQ--templatestyles-0000004F-QINU`"'<cite class="citation cs2">''authors''</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=authors&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1%2FArchive+18" class="Z3988"></span> <span class="cs1-visible-error citation-comment"><code class="cs1-code">{{[[Template:citation|citation]]}}</code>: </span><span class="cs1-visible-error citation-comment">Unknown parameter <code class="cs1-code">&#124;authors=</code> ignored ([[Help:CS1 errors#parameter_ignored|help]])</span>

authors {{citation}}: Unknown parameter |authors= ignored (help)

{{cite book |title=authors |authors=Last FM, Last FM |ref=harv}}

'"`UNIQ--templatestyles-00000053-QINU`"'<cite class="citation book cs1">''authors''.</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=authors&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1%2FArchive+18" class="Z3988"></span> <span class="cs1-visible-error citation-comment"><code class="cs1-code">{{[[Template:cite book|cite book]]}}</code>: </span><span class="cs1-visible-error citation-comment">Invalid <code class="cs1-code">&#124;ref=harv</code> ([[Help:CS1 errors#invalid_param_val|help]])</span>; <span class="cs1-visible-error citation-comment">Unknown parameter <code class="cs1-code">&#124;authors=</code> ignored ([[Help:CS1 errors#parameter_ignored|help]])</span>

authors. {{cite book}}: Invalid |ref=harv (help); Unknown parameter |authors= ignored (help)

But, there is a bug that causes it to do so when the template has a date:

{{citation |title=authors |authors=Last FM, Last FM |year=1995}}

'"`UNIQ--templatestyles-00000057-QINU`"'<cite class="citation cs2">''authors'', 1995</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=authors&rft.date=1995&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1%2FArchive+18" class="Z3988"></span> <span class="cs1-visible-error citation-comment"><code class="cs1-code">{{[[Template:citation|citation]]}}</code>: </span><span class="cs1-visible-error citation-comment">Unknown parameter <code class="cs1-code">&#124;authors=</code> ignored ([[Help:CS1 errors#parameter_ignored|help]])</span>

authors, 1995 {{citation}}: Unknown parameter |authors= ignored (help)

{{cite book |title=authors |authors=Last FM, Last FM |ref=harv |year=1995}}

'"`UNIQ--templatestyles-0000005B-QINU`"'<cite class="citation book cs1">''authors''. 1995.</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=authors&rft.date=1995&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1%2FArchive+18" class="Z3988"></span> <span class="cs1-visible-error citation-comment"><code class="cs1-code">{{[[Template:cite book|cite book]]}}</code>: </span><span class="cs1-visible-error citation-comment">Invalid <code class="cs1-code">&#124;ref=harv</code> ([[Help:CS1 errors#invalid_param_val|help]])</span>; <span class="cs1-visible-error citation-comment">Unknown parameter <code class="cs1-code">&#124;authors=</code> ignored ([[Help:CS1 errors#parameter_ignored|help]])</span>

authors. 1995. {{cite book}}: Invalid |ref=harv (help); Unknown parameter |authors= ignored (help)

I have fixed the bug in the sandbox:

{{citation/new |title=authors |authors=Last FM, Last FM |year=1995}}

'"`UNIQ--templatestyles-0000005F-QINU`"'<cite class="citation cs2">''authors'', 1995</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=authors&rft.date=1995&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1%2FArchive+18" class="Z3988"></span> <span class="cs1-visible-error citation-comment"><code class="cs1-code">{{[[Template:citation|citation]]}}</code>: </span><span class="cs1-visible-error citation-comment">Unknown parameter <code class="cs1-code">&#124;authors=</code> ignored ([[Help:CS1 errors#parameter_ignored|help]])</span>

authors, 1995 {{citation}}: Unknown parameter |authors= ignored (help)

{{cite book/new |title=authors |authors=Last FM, Last FM |ref=harv |year=1995}}

'"`UNIQ--templatestyles-00000063-QINU`"'<cite class="citation book cs1">''authors''. 1995.</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=authors&rft.date=1995&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1%2FArchive+18" class="Z3988"></span> <span class="cs1-visible-error citation-comment"><code class="cs1-code">{{[[Template:cite book|cite book]]}}</code>: </span><span class="cs1-visible-error citation-comment">Invalid <code class="cs1-code">&#124;ref=harv</code> ([[Help:CS1 errors#invalid_param_val|help]])</span>; <span class="cs1-visible-error citation-comment">Unknown parameter <code class="cs1-code">&#124;authors=</code> ignored ([[Help:CS1 errors#parameter_ignored|help]])</span>

authors. 1995. {{cite book}}: Invalid |ref=harv (help); Unknown parameter |authors= ignored (help)

—Trappist the monk (talk) 00:05, 25 May 2016 (UTC)

Looks good to me. I checked the documentation to see if |authors= was mentioned in the harv/CITEREF section, and I did not find it, so I don't think this will introduce any changes that contravene the documentation. – Jonesey95 (talk) 14:44, 25 May 2016 (UTC)

Good. Thank you.

—Trappist the monk (talk) 15:39, 25 May 2016 (UTC)

An observation:

* |ref=harv: Creates an anchor of the format CITEREFauthorslastnameyear suitable for a {{harv}}, {{sfn}} etc.
— H:CS1#Anchors

Change CITEREFauthorslastnameyear to CITEREFauthorlastnameyear or CITEREFauthor(s)lastnameyear

65.88.88.127 (talk) 18:36, 31 May 2016 (UTC)

The new code that renders date ranges with an endash when the template-source uses a hyphen separator, revealed this bug where a date value beginning with 'n.d.' is accepted regardless of any additional stuff that may be in the parameter value. That is remedied in the sandbox:

Cite book comparison

Wikitext	{{cite book|date=n.d. (1814-1818)|last=Anonymous|location=Birmingham|publisher=H. Wadsworth|title=The Holly & Ivy|url=http://www.vwml.org/record/CJS1/10/2/432/1}}
Live	Anonymous (n.d. (1814-1818)). The Holly & Ivy. Birmingham: H. Wadsworth. {{cite book}}: Check date values in: |date= (help)
Sandbox	Anonymous (n.d. (1814-1818)). The Holly & Ivy. Birmingham: H. Wadsworth. {{cite book}}: Check date values in: |date= (help)

The 'nd' version does not suffer the same problem:

Cite book comparison

Wikitext	{{cite book|date=nd (1814-1818)|last=Anonymous|location=Birmingham|publisher=H. Wadsworth|title=The Holly & Ivy|url=http://www.vwml.org/record/CJS1/10/2/432/1}}
Live	Anonymous (nd (1814-1818)). The Holly & Ivy. Birmingham: H. Wadsworth. {{cite book}}: Check date values in: |date= (help)
Sandbox	Anonymous (nd (1814-1818)). The Holly & Ivy. Birmingham: H. Wadsworth. {{cite book}}: Check date values in: |date= (help)

—Trappist the monk (talk) 19:55, 4 June 2016 (UTC)

It steals has he same sample date, and it's been a week. Eurocus47 (talk) 17:36, 4 June 2016 (UTC)

@Eurocus47: What is "this page"? There is no mention of a "sample date" either at Help:Citation Style 1 or at Help talk:Citation Style 1, other than in ths section. --Redrose64 (talk) 11:38, 5 June 2016 (UTC)

Following up on this conversation, I have added support for generational suffixes in Vancouver style when cs1|2 templates use |lastn= / |firstn= with |name-list-format=vanc or use |vauthors=:

using |vauthors=:

Last FM Sr. pass.

Last FM Jr. pass.

Last FM 2nd. pass.

Last FM 3rd. pass.

Last FM 4th. pass.

using |lastn= / |firstn= with |name-list-format=vanc

Last FM Sr. pass.

Last FM 9th. pass.{{cite book}}: CS1 maint: numeric names: authors list (link)

The code accepts ordinals 2nd–9th.

As part of this change, I've tweaked the Vancouver error messaging a bit so that error messages attempt to indicate why something is not right:

when suffix is out of range

Last FM 10th. fail: L/F suffix out of range. {{cite book}}: Vancouver style error: suffix in name 1 (help)CS1 maint: numeric names: authors list (link)

Last FM 22nd. fail: suffix out of range. {{cite book}}: Vancouver style error: suffix in name 1 (help) – 22nd not valid but the code can't tell if FM is part of a multi-word 'last' name or first & middle initials because there is no last name mixed case requirement

too many initials

Last F 4th. fail: L/F too many initials w/suffix.{{cite book}}: CS1 maint: numeric names: authors list (link)– an error but not an error; could be a case where |first= has a name that is written in all uppercase

Last F. fail: L/F too many initials.– an error but not an error; same as above

Last F 4th. fail: too many initials w/suffix. {{cite book}}: Vancouver style error: initials in name 1 (help) – |vauthors= too many uppercase initials is an error

Last F. fail: too many initials. {{cite book}}: Vancouver style error: initials in name 1 (help)

space between initials:

Last F M. fail: space between initials. {{cite book}}: Vancouver style error: initials in name 1 (help) – |vauthors=Last F M

Last F M Jr. fail: space between initials w/suffix. – |vauthors=Last F M Jr – not detected as an error becuase module doesn't (shouldn't) test for name length because some name parts are one character

last name and suffix without initials (|vauthors= only):

Last Jr. fail: no initials. {{cite book}}: Vancouver style error: name in name 1 (help) – |vauthors=Last Jr – because Jr is valid suffix, module sees 'Last' as what should be the name's initials

extraneous punctuation characters dot and semicolon:

Last F Sr. fail: punctuation. {{cite book}}: Vancouver style error: punctuation in name 1 (help) – |vauthors=Last FM Sr. – a very common error especially when name lists are copied and pasted from a PMID page

non-Latin characters:

Last FM, 3rd. fail: improper punctuation. {{cite book}}: Vancouver style error: non-Latin character in name 2 (help) – |vauthors=Last FM, 3rd – because the comma indicates the start of a new name; '3' is not a letter

Last ЕМ 3rd. fail: non-Latin characters. {{cite book}}: Vancouver style error: non-Latin character in name 1 (help) – |vauthors=Last ЕМ 1st – Cyrillic characters in this example which look remarkably like Latin characters

wikilinks:

Lincoln A. fail: wikilink. – |vauthors=[[Abraham Lincoln|Lincoln A]]

missing comma in |vauthors= name list:

Last F Last M. fail: missing comma. {{cite book}}: Vancouver style error: missing comma in name 1 (help) – |vauthors=Last F Last M

Not altogether successful but not a complete failure either. What I think this does mean is that the whole name-handling is a mess. I've been wondering if all of the name-handling code shouldn't be moved out of Module:Citation/CS1 into its own page and be given a good rethink with the goal of eliminating redundancies and improving (especially for Vancouver style) error handling, detection, and reporting; metadata support; add a semantically correct |vauthor= when a template lists only one name in that format; perhaps other stuff. Perhaps after the next update.

—Trappist the monk (talk) 11:27, 23 May 2016 (UTC)

Thanks for implementing this!

I found a bug ~

{{cite journal | vauthors = Pometto AL 3rd, Crawford DL, Last FM | title = Blah}}

appends the "3rd" to each subsequent author:

Pometto AL 3rd, Crawford DL, Last FM. "Blah". {{cite journal}}: Cite journal requires |journal= (help)

Moving the "3rd" from the first author to the second author:

{{cite journal | vauthors = Pometto AL, Crawford DL 3rd, Last FM | title = Blah}}

Pometto AL, Crawford DL 3rd, Last FM. "Blah". {{cite journal}}: Cite journal requires |journal= (help)

  ~ Tom.Reding (talk ⋅dgaf)  00:11, 6 June 2016 (UTC)

Fixed in the live module (I had more difficulties copying the sandbox to live than I had fixing the vauthors bug).

—Trappist the monk (talk) 00:35, 6 June 2016 (UTC)

Something has gone wrong with {{citation}}, {{cite journal}}, etc., in the last few minutes. Whenever I use them, I get a big red error message "Lua error in Module:Citation/CS1 at line 3231: attempt to index global 'cs1' (a nil value)." —David Eppstein (talk) 00:30, 6 June 2016 (UTC)

A temporary glitch? Seems to be ok again. —David Eppstein (talk) 00:32, 6 June 2016 (UTC)

Yes, a simple |vauthors= bug fix turned ugly.

—Trappist the monk (talk) 00:58, 6 June 2016 (UTC)

Both |authors= and |last= correctly give the 'multiple-author-lists' error when |vauthors= is present, but |first= does not ~

Using |authors=:

{{cite journal|vauthors=Lastone FM, Lasttwo FM|author=Lastoneauthor FM|title=Blah}}

Lastonetest FM. "Blah". {{cite journal}}: Cite journal requires |journal= (help); More than one of author-name-list parameters specified (help)

Using |last=:

{{cite journal|vauthors=Lastone FM, Lasttwo FM|last=Lastonelast FM|title=Blah}}

Lastonelast FM. "Blah". {{cite journal}}: Cite journal requires |journal= (help); More than one of author-name-list parameters specified (help)

Using |first=:

{{cite journal|vauthors=Lastone FM, Lasttwo FM|first=Lastonefirst FM|title=Blah}}

"Blah". {{cite journal}}: |first= missing |last= (help); Cite journal requires |journal= (help); More than one of author-name-list parameters specified (help)

The same is true for |editor-first=.   ~ Tom.Reding (talk ⋅dgaf)  14:05, 6 June 2016 (UTC)

What I would guess is happening is that the |first= without |last= test should be displayed also, but is likely relying on the same bit of code that identifies the duplication in |last= and |vauthors=. --Izno (talk) 14:18, 6 June 2016 (UTC)

The module first looks for |last= (and aliases). If none are found, the module looks for |vauthors= and finally |authors=. |first= is not examined. There is only one list type rendered according to the priority order just described.

I've tweaked the sandbox so that it will look for |firstn= where n is one of the set [nil, 1, 2]. The previous examples using the sandbox:

{{cite journal/new|vauthors=Lastone FM, Lasttwo FM|author=Lastoneauthor FM|title=Blah}}

Lastonetest FM. "Blah". {{cite journal}}: Cite journal requires |journal= (help); More than one of author-name-list parameters specified (help)

{{cite journal/new|vauthors=Lastone FM, Lasttwo FM|last=Lastonelast FM|title=Blah}}

Lastonelast FM. "Blah". {{cite journal}}: Cite journal requires |journal= (help); More than one of author-name-list parameters specified (help)

{{cite journal/new|vauthors=Lastone FM, Lasttwo FM|first=Lastonefirst FM|title=Blah}}

"Blah". {{cite journal}}: |first= missing |last= (help); Cite journal requires |journal= (help); More than one of author-name-list parameters specified (help)

—Trappist the monk (talk) 15:32, 6 June 2016 (UTC)

Are there any plans to extend the multi-author error checks for only 2 authors? I'm guessing this hasn't been done yet in order to avoid misidentifying single authors like |author=Charles II, William W as 2 authors. Can the current multi-author error check be made to only flag single-comma author/editor fields which don't contain Roman-numeral-like initials [IV]+, but which otherwise follow Vancouver format? I'm currently migrating these cases (where appropriate) to |vauthors=, but I only happen on them if they populate Category:CS1 maint: Multiple names: authors list by some other means.

Currently no 2-author error:

{{cite journal|author=Last FM, Last FM|title=Blah}}

Last FM, Last FM. "Blah". {{cite journal}}: Cite journal requires |journal= (help)

  ~ Tom.Reding (talk ⋅dgaf)  20:30, 6 June 2016 (UTC)

In the article New World wine the book by José del Pozo Historia del vino chileno is cited two times: one time pages 24–34 and the other 35–45. Is there a way to cite it only one time specifying different pages intervals, or should I use {{cite book}} two times separately?--Carnby (talk) 15:39, 8 June 2016 (UTC)

<ref>{{harvnb}}</ref> or {{sfn}} in place of the <ref>{{cite book}}</ref> references; if necessary create a bibliography section and place the {{cite book}} template there (without page number information which properly belongs in the short-form templates); and Bob's yer Uncle.

—Trappist the monk (talk) 15:52, 8 June 2016 (UTC)