Template talk:Citation/core/Archive 13

AuthorType
There are occasions where it is useful to specify the author type, such as illustrator, developer, writer, etc. I have added AuthorType, EditorType and OtherType to the sandbox— these place the type in parenthesis immediately after the author.

Full sample:

Example:

---— Gadget850 (Ed)  talk 13:58, 18 October 2011 (UTC)
 * Oppose. Every little crufty change like this that we make seems harmless in itself, but combined they make the template very hard to change and they add a lot of load to the Wikipedia servers. I don't think the usefulness of this one (which I agree is nonzero) outweighs its cruftiness. —David Eppstein (talk) 15:26, 18 October 2011 (UTC)


 * Oh well. I had hoped to update some citation templates that don't use citation/core, but I will just have to drop that dream. ---— Gadget850 (Ed)  talk 16:02, 18 October 2011 (UTC)


 * This departs from customary citation practice outside Wikipedia in two ways; it encourages listing of contributors who are not traditionally listed in a bibliography, and does not necessarily group the same kind of contributors together. Also, it does not provide an opportunity to save space by a word such as "illustrator" only once in the listing.


 * Further, it creates two different labeling systems for different ill-defined overlapping groups. The first group is authors, editors, and others. The second group is the collection of everything the Wikipedia editor decides to type in any of the new fields. This vagueness and overlap will prevent Wikipedia editors from forming a clear mental model of what the citation templates are all about. Jc3s5h (talk) 16:13, 18 October 2011 (UTC)

I see consensus is against this, so I'm abandoning this. Folks will just have to continue mixing citation templates that use other styles or abusing the Given1 field. As an exercise, I did check performance issues. I tested a page containing 1,000 uses of citation/core: current use, sandbox use withour invoking AuthorType and sandbox use invoking AuthorType:

Current version
 * Preprocessor node count: 252001/1000000
 * Post-expand include size: 1150200/2048000 bytes
 * Template argument size: 189000/2048000 bytes
 * Expensive parser function count: 0/500

Sandbox not using AuthorType Preprocessor node count: 253801/1000000 Post-expand include size: 1141200/2048000 bytes Template argument size: 187200/2048000 bytes Expensive parser function count: 0/500

Sandbox using AuthorType
 * Preprocessor node count: 254701/1000000
 * Post-expand include size: 1115100/2048000 bytes
 * Template argument size: 172800/2048000 bytes
 * Expensive parser function count: 0/500

Difference between current and sandbox using AuthorType
 * Preprocessor node count: 2700 bytes difference per 1000; 2 bytes unused each; 3 bytes per use each
 * Post-expand include size: 35,100 bytes difference per 1000; 9 bytes unused each; 35 bytes per use each
 * Template argument size: 16,200 bytes difference per 1000; 16 bytes unused each; 17 bytes per use each

---— Gadget850 (Ed)  talk 15:25, 19 October 2011 (UTC)

Closed. Cite video will continue to abuse the author fields. Closing consideration of updating cite video game, cite episode, cite serial, cite comic, comic strip reference or similar templates. ---— Gadget850 (Ed)  talk 12:37, 2 November 2011 (UTC)

Who stuffed up the positioning of the Chapter parameter for book class objects?


displays as



when it should display as


 * Jonesson, Jones "Jonesson's life" Academic, Susan ed. A book of Joneses

This causes false attribution of the work cited and is a critical error in terms of Verification. Fifelfoo (talk) 00:09, 10 November 2011 (UTC)

HRRM. Allow me to copy and paste. Fifelfoo (talk) 00:17, 10 November 2011 (UTC)

My fault; just realised why. Fifelfoo (talk) 00:18, 10 November 2011 (UTC)

Translators (again)
I would like to propose an additional field which allows editors to credit one or several translators, as per Template talk:Cite book/Archive 9. This has proposal has already been raised and discussed several times on that talk page and I fail to understand why it has not been implemented yet. Is there a technical problem, or are there any other objections? --Florian Blaschke (talk) 13:48, 23 November 2011 (UTC)
 * There are those who object to the use of citation templates principally on the grounds that they take a relatively long time to render. Each additional parameter increases the rendering time. We should perhaps be looking to decrease, rather than increase, the number of parameters, in order to improve performance. This template already provides the parameter Other, used by e.g. - see its documentation for others. -- Red rose64 (talk) 14:52, 23 November 2011 (UTC)
 * We could reduce the rendering time if we removed the COinS tag pieces. I know there are a few obscure pieces of software that use this (Zotero?), but is it really worth the cost? We probably render literally millions of COinS tags a day, and maybe one or two get used. Kaldari (talk) 02:16, 24 November 2011 (UTC)

To return to the original question: Proposing this at cite book is fine, but it cannot be fixed there. This is the correct place. The addition of translator fields would be fairly trivial (simply copy the editor markup and rename). Please make your case here for why this would be valuable---— Gadget850 (Ed)  talk 13:47, 24 November 2011 (UTC)

Editors are hardwired to have "et al" after the third.
I'm wanting to display the fourth editor of a book for the article Experimental political science, because that editor has a Wikipedia page. However, I get "et al" after the name of the third editor. I suspect this is because the code in Citation (used by Cite book) is: #if: |&amp;#32;et al.

Could something please be done about making this more symmetrical with the author practice (with Trunc or another interface, although the latter would require modifications to the cite book|web|etc templates...)? Thanks! Allens (talk) 18:49, 27 November 2011 (UTC)
 * You are correct in that the issue is here. By design, the fourth editor never shows and is replaced by et. al. ---— Gadget850 (Ed)  talk 19:33, 27 November 2011 (UTC)
 * Why by design is it not flexible (and not defaulting to the same number as authors, particularly when no authors are cited)? I also note the lack of documentation of this, if it is meant to be the case. Allens (talk) 20:56, 27 November 2011 (UTC)
 * It is documented on this template pages— see EditorSurname2, documentation on the other templates is not consistent. I will look at this in a bit. Do you think there would be any need for more editors? ---— Gadget850 (Ed)  talk 23:12, 27 November 2011 (UTC)
 * In terms of documentation, I'm not finding EditorSurname2 in the template doc-page, nor does a search for "et al" find anything about it; ? If the fourth or more editor has a Wikipedia page, it would be nice to be able to link to it easily, although fortunately the editor in question in Experimental political science is also the author of an article referenced. I'll try to take a look to see if any "official" citation formats call for differing numbers before "et al" of editors as opposed to authors. Thanks for taking a look at the documentation problem... Allens (talk) 02:05, 29 November 2011 (UTC)
 * Not sure which documentation page you're looking at, but since this page is Template talk:Citation/core, the documentation to look at is that shown at Template:Citation/core, where we find under Template:Citation/core, the following:
 * EditorSurname1 first editor's surname or last name.
 * EditorSurname2, EditorSurname3, EditorSurname4 second, third, and fourth editors’ surname or last name. The fourth is not actually used, but causes et al. to be generated.
 * it's in approximately alphabetic order. -- Red rose64 (talk) 14:50, 29 November 2011 (UTC)
 * Ahh... I see the confusion. I was looking at Template:Citation's documentation (which is where users of it are going to be looking, most likely, not at the core one...) Allens (talk) 15:01, 29 November 2011 (UTC)
 * The documentation at Template:Citation is lacking in quite a few areas, because the template is used for a wide variety of different types of source. Some parameters have different effects depending upon the presence or absence of other parameters, and trying to make the doc cover all possibilities is a nightmare. -- Red rose64 (talk) 16:11, 29 November 2011 (UTC)

It used to be that authors were hardwired to three before "et al" kicked in. I guess when we changed that we didn't also change editor to match. I wouldn't mind at least allowing for the possibility of more editors. —David Eppstein (talk) 02:07, 29 November 2011 (UTC)


 * While I am in this, editors don't honor NameSep— they always use a comma. That needs to match the author names. ---— Gadget850 (Ed)  talk 03:15, 29 November 2011 (UTC)

OK, now in sandbox:
 * Supports up to eight editors
 * Et. al controlled by EditorTrunc, defaults to 3
 * Supports NameSep
 * Supports EditorMask

Current templates such as cite book or citation will need to be updated to pass these parameters.

current









sandbox













---— Gadget850 (Ed)  talk 16:20, 29 November 2011 (UTC)


 * Excellent! Thanks! Should some examples be put into Template:Citation/core/testcases? (I note that ArchiveURL and URL are having a problem in the "full citation" test cases, unless that's deliberate...) Allens (talk) 22:08, 30 November 2011 (UTC)


 * The ArchiveURL and URL appear to be tests using the current template and do not reflect any issue on the sandbox version. Please note that the editor markup is in two chunks (no author, editor and author, editor) and I still have to do the second, so this does not work yet if there is an author value. I am looking to see if that could be simplified, but that will probably wait till later. ---— Gadget850 (Ed)  talk 22:29, 30 November 2011 (UTC)





Duplicated period
❌

One of the complaints I have encountered is duplicated periods, especially on first names with initials:

The only current solution is to remove the trailing period from the field, but this means that it is missing from the CoinS metadata.

I created User:Gadget850/Remove trailing period as a fix to this:

I have tested this, and it removes the period for a string of up to 500 characters. If we implement this, it should apply only to first names. I have seen complaints about the publisher field, but this is usually because the corporate designation such as Inc. was included, which the guidelines and style guides advise against. ---— Gadget850 (Ed)  talk 14:41, 10 October 2011 (UTC)


 * Has the fix been tested for performance on a page with many citations? Jc3s5h (talk) 15:12, 10 October 2011 (UTC)


 * Good question. I just substed the template markup it one level to eliminate some transclusions. ---— Gadget850 (Ed)  talk 15:16, 10 October 2011 (UTC)


 * Looks like each use adds:
 * Preprocessor node count: 419 bytes
 * Post-expand include size: 1999 bytes
 * Template argument size: 528 bytes
 * Expensive parser function count: 0
 * PageSpeed shows 94/100 for both test pages of 100 citations. I am open to any optimization here.---— Gadget850 (Ed)  talk 17:43, 10 October 2011 (UTC) ---—  Gadget850 (Ed)  talk 17:43, 10 October 2011 (UTC)


 * Actually, this is only an issue when there is no date; see above. If the date exists, then there is no need to invoke the period remover. ---— Gadget850 (Ed)  talk 12:59, 12 October 2011 (UTC)

There is some reason to think cite book was based on APA style. In APA, when one can't determine a date for a work, the date is listed as "n.d." However, in Wikipedia the situation is usually that the editor didn't bother to give the date, rather than it being hard to discover the date. So automatically putting in "(n.d.)." whenever all the date-related parameters are missing is probably a bad idea. Jc3s5h (talk) 14:20, 12 October 2011 (UTC)


 * And I am not proposing that. Simple logic such as  would do the trick. Thus, the remover would only be invoked if there is no date and that bit of overhead is not added. But it would have to check for date, year and month fields. This is probably too complicated, but I will have to think on it a bit.
 * Might be easier to simply add an author-terminator parameter that would simply not include the period. ---— Gadget850 (Ed)  talk 15:26, 12 October 2011 (UTC)


 * Definitely easier to add author-terminator. The author link markup makes it very hard to get a clean hook into it. The template was a good exercise and I will probably develop it into general use. ---— Gadget850 (Ed)  talk 23:25, 12 October 2011 (UTC)


 * A better but probably harder solution is to not remove the period from the data, but from the markup around the data, if this will have any effect at all on the COINS metadata, which we should not be falsifying for the sake of formatting. — SMcCandlish  Talk⇒ ʕ(Õلō)ˀ  Contribs. 19:47, 16 December 2011 (UTC)

Backwards specification and ownership behaviour over Template:Cite report
Cite report was heavily expanded in 2009 to deal with unpublished material, specifically, reports which were "unpublished" in the citation sense of never being offered for sale; but, verifiable in the wikipedia sense of published as available for consultation, and uttered by responsible authorities.

Gadget850 improved Cite report to use Citation/core in September 2011, incidentally breaking fundamentally behaviour dealing with the |title= parameter; behaviour following Turabian Reference List style and Harvard of unpublished works taking neither italics not quote marks.

Help:Citation Style 1 isn't a useful specification here: it is backwards specified from published source practice. Would editors discuss the appropriate specification for the title of unpublished works on wikipedia, in particular, reports such as which were reliably uttered but not published in the same sense as a webpage, book, newspaper article or journal article. Fifelfoo (talk) 12:30, 30 November 2011 (UTC)
 * Reasons to support non-italic and non-quote behaviour as the specification:
 * Turabian reference list style
 * Harvard style
 * Successful behaviour on wikipedia over 2009–2011 with wide adoption for a rare citation template
 * Unpublished (by citation standards) materials not being readily located by title anyway
 * That italics are used in most citation styles for independently published works, ie books, journals or newspapers, whereas quotes are used for contained elements of a greater work, ie chapters etc. Unpublished reports, PhD theses, etc, aren't independent published works due to lack of publication.  Fifelfoo (talk) 12:40, 30 November 2011 (UTC)


 * You won the argument and you are bitching about it? You want a template for unpublished reports, you got it. You want a template that is not like the others, you got it. Everything is back to where it was before I got involved, so I certainly don't own anything. Have a nice day. And why are you discussing it here? ---— Gadget850 (Ed)  talk 12:49, 30 November 2011 (UTC)
 * "I won"? You seem to believe that 3 years of a template existing in the display manner of other Cite x styles means that it doesn't belong in the family?  This is about whether reports should be specified to be cited in the "Cite x" house style of wikipedia with italicised titles or not.  There isn't winning, there is establishing how the Cite series of templates treats unpublished reports.  Since 2009 that family has treated them without italics for titles.  This discussion can settle the specification, and then people can bother to implement after a specification has been created as the result of consensus.  (I'd argue the consensus from 2009 is indicative). Fifelfoo (talk) 12:53, 30 November 2011 (UTC)
 * Hmmm... This is probably not the optimum place to raise this, but I have been grumping on and off for a while about the (lack of a) template engineering process. This is not a big deal for low usage templates, but it is a big deal for high usage templates. The remarks immediately above speak of a "house style" for WP cites -- there is no such thing. However, there is a de-facto style for templated cites. That de-facto style is not documented anywhere. With that as an intro, I'll make a proposal for the citation and templates: Let's create an article for each of those templates like Template:template_name/requirements_specification, and lets create an environment enforced by the WP editing community that templates having such a "requirement specification" article must comply with those requirements. Let's sort out disagreements about citation requirements on those requirement spec articles. Lots of other stuff flows from that, but that's as far as I want to take it now. Comments? Wtmitchell  (talk) (earlier Boracay Bill) 13:56, 30 November 2011 (UTC)
 * We do not have preferred styles, but we do have a series of well used templates based on citation/core that share the same styles. I had the same observation on a lack of common usage instructions, had some discussion, made a proposal and started Help:Citation Style 1. I suppose you could call it a specification, but the intent is to describe how current templates work. In doing so, I have identified those templates using the CS1 style.
 * I have also updated templates that are similar but not fully compliant with CS1. I have never implemented anything other than very minor changes without discussion.
 * Citation/core formats titles per the Chicago style: long works in italics and short works in quotes. Cite report originally had no format for titles. I did a sandbox update and made a proposal on the talk page; after a reasonable period, I made the changes per WP:SILENCE. I formatted the title in quotes; Fifelfoo objected to the title format stating "the update to cite core broke the non-italicisation of the titles of unpublished works." We had a discussion that I could quickly see was not going to reach any consensus and I reverted to the original markup. No CS1 template uses unformatted titles and no CS1 templates are for unpublished works. If there is consensus that templates should format unpublished works in this manner, then we need to add that functionality to citation/core instead of making an HTML hack. ---— Gadget850 (Ed)  talk 15:07, 30 November 2011 (UTC)
 * I think Gadget850's work in unifying Citation 1, and documenting it is excellent. I agree that the implementation shouldn't be the kludge that I forced in a pique.  I'm glad that we've laid out the issue to seek advice on which behaviour to specify, then implement, and document.  I think Wtmitchell's suggestion that we should formally specify CS1 is a good one.  One reason we ought to specify CS1, as well as documenting it, is that CS1 has become a major international citation style simply by widespread adoption by wikipedia users.  I'm happy with whatever specification editors make consensus on regarding unpublished reports.  Fifelfoo (talk) 23:00, 30 November 2011 (UTC)

This needs to be broken into two issues:
 * Should the title be plain formatted? Discussion should be at Manual of Style/Titles.
 * Should we cite unpublished works? Discussion should be at Identifying reliable sources.

I discourage the use of Cite x or Cite xxx, as they are meaningless. Not every template with a name beginning with cite uses the same style as CS1 (Cite newspaper The Times for example), nor are they all citation templates (cite quote for example). Unfortunately, the casual editor is not aware of stylistic difference and most probably think that all the cite templates were created under some master plan. ---— Gadget850 (Ed)  talk 13:14, 1 December 2011 (UTC)
 * MOS TItles covers body text—MOS Titles doesn't cover citations, as both CITEVAR and the wide variety of manual citations in Science articles where titles aren't italicised demonstrates. We've been citing unpublished material for years, if they're published in the term-of-art sense of WP:V of having been uttered by a body that takes responsibility for them, and are available for to be verified against; they also need to meet the secondary/independent-of-the-subject criteria.
 * Cite report was designed specifically to match the most common other Cite x format citations, which have become the CS1 set. Fifelfoo (talk) 13:19, 1 December 2011 (UTC)


 * CS1 is a mishmash of styles. Titles use the Chicago style of long works in italics and short works in quotes. Per Chicago:


 * 8.184 TITLES OF UNPUBLISHED WORKS


 * Titles of unpublished works--theses, dissertations, manuscripts in collections, unpublished transcripts of speeches, and so on--are set in roman type, capitalized as titles, and enclosed in quotation marks. Names of manuscript collections take no quotation marks.


 * ---— Gadget850 (Ed)  talk 13:28, 5 December 2011 (UTC)
 * This is a very cogent argument: that CS1 style uses Chicago style in relation to titles. My only counter argument would be that Cite report was built to match other Cite x styles that have become CS1; and, that Cite report's behaviour has operated successfully from 2009 to 2011.  I guess I'm at the stage where I'd like a third user or third users to close this stylistic debate so we can have a happy implementation.  Gadget850, would you support having a specification separate from implementation and documentation for CS1?  I'd be happy to work on that with you. Fifelfoo (talk) 01:37, 7 December 2011 (UTC)


 * In my view, Turabian is derivative, "renegade" in many respects, and simply not authoritative; in over 20 years in and out of academic environments, I have never once had a professor in an field require Turabian style, have had many forbid it, and had most never mention it, any more than they mention Bedford Guide style or other minority citation styles. Harvard style is of limited (and from what I can tell, decreasing) favor and influence, and is not well liked on Wikipedia generally, outside of articles on particular disciplines. I would not want to use either of these as authorities for what to do on Wikipedia. While I have my issues with Chicago on various points, it's influence on our citation templates (and on the off-WP academic and professional publishing world) is strong enough that I have to side strongly with it's suggestion of putting the titles of unpublished works in quotation marks, in the rare cases that this actually comes up here. Putting them in  italics nor quotation marks is out of the question, as it will confuse readers and editors, lead to editwarring over "correcting" the citations, etc., etc.
 * Some of Fifelfoo's arguments in favor of using neither aren't convincing. "Successful behaviour on wikipedia over 2009–2011 with wide adoption for a rare citation template" is meaningless, since changing the output of the template in this minor way won't break this "successful behavior"; the fact that the template is rare is precisely why it took a long time for any to notice and object to its weirdness. "Unpublished (by citation standards) materials not being readily located by title anyway" is irrelevant; it's not about locating the resource, but even understanding the citation; people are going to think "WTF is this?" when they seem something with neither italics or quotes, and "fix" it one way or the other. "That italics are used in most citation styles for independently published works, ie books, journals or newspapers, whereas quotes are used for contained elements of a greater work, ie chapters etc. Unpublished reports, PhD theses, etc, aren't independent published works due to lack of publication." Right, so this is argument for quotation marks, which are also used for minor works that are  part of a greater work, such as songs (in the sheet music sense; yes, of course, albums are greater works that do contain songs). It's not an argument for no formatting of the title at all.
 * So: Use quotation marks per Chicago and our fairly consistent reliance on it in this sphere, and per our reader/editors' needs and expectations. Alternatively, use italics or quotation marks as one would if the work were published (and maybe have a parameter that switches them or something) if people are going to whine and cry about it, but don't use neither. — SMcCandlish   Talk⇒ ʕ(Õلō)ˀ  Contribs. 20:12, 16 December 2011 (UTC)