Template talk:Citation/Archive 7

RFC: Same rules for CS1 and Citation
As a side-effect of using Module:Citation/CS1 to render the Citation template, all the warning messages issued for Citation Style 1 will now be issued for Citation. (Many of these warning messages are not turned on by default yet.) This means that editors who use the Citation template will have to consult Help:Citation Style 1 to determine the acceptable parameter values. Does the user community ratify this change? Jc3s5h (talk) 17:24, 30 April 2014 (UTC)

Same rules for CS1 and Citation:Discussion
Comment. A possible reason to have different rules for Citation and Citation Style 1 is that Citation only lends itself to footnote citations; comma separators between elements are not traditional for bibliographies (also known as reference lists or lists of works cited). The most popular footnote citation style is Chicago Manual of Style Documentation One style. Citation Style 1 more closely resembles APA style. The community may wish to emulate different printed style manuals for Citation vs. Citation Style 1.

To see all the error messages, view the directions at Help:CS1 errors

If the rules are made the same, the documentation will require restructuring. Jc3s5h (talk) 17:24, 30 April 2014 (UTC)
 * +1 &rarr;Jc3s5h --User:Ceyockey ( talk to me ) 00:29, 2 May 2014 (UTC)


 * It seems to me that those of us not steeped in citation wizardry have no idea what the practical implications of this might be. Could you illustrate with some examples? I know this might be a lot of work but it sounds like this is something which could have wide-ranging effects, and even touch off violent protests, so it better be well-ventilated. EEng (talk) 17:38, 30 April 2014 (UTC)
 * I haven't had time to review the parameter descriptions between CS1 and Citation. However, I did look through the possible error messages at CS1 errors. I found that some of the errors that might occur are things that really ought to be fixed, not just a matter of preference. The preference items I found were date related. One is the CS1 requirement that the accessdate parameter should not be present unless there is a URL; this requirement is equivocal in the Citation documentation, and there has been discussion in various places about whether this requirement is a good idea. The other preference issue is date format. CS1 adopts Manual of Style/Dates and numbers, Dates and years section, with certain exceptions because some date formats that could be used in general are not applicable to citations, and because the technology is unable to deal with certain dates (for example, 29 February 1600 in England). The Citation template does not have any date format requirements other than consistency within an article. Jc3s5h (talk) 21:25, 30 April 2014 (UTC)


 * Some basic parameter name usage is quite different in CS1 vs cite. For instance, to cite a conference paper in CS1, one should use cite conference with the paper's title in title and the conference name in conference. To do so in citation, one should use citation with the paper's title in contribution and the conference name in title. Given that level of basic incompatibility, it seems premature to direct users of citation to documentation that will tell them the wrong parameter names. —David Eppstein (talk) 17:50, 30 April 2014 (UTC)
 * This can be fixed in the documentation with in includeonly code that checks the pagename and gives slight different instructions for different templates. Another solution is to normalize citation's parameters to those of CS1, and fix extant transclusions, but that might be a lot of trouble.  — SMcCandlish ☺ ☏ ¢ ≽ʌⱷ҅ᴥⱷʌ≼  03:51, 1 May 2014 (UTC)
 * That would make more sense if the CS1 parameters were anything close to normalized themselves. Or do you think it's sane that citing a chapter in a single-author book, a chapter in an edited volume, and a chapter in a conference proceedings are all so different from each other in CS1? —David Eppstein (talk) 04:38, 1 May 2014 (UTC)


 * Information: and the CS1 templates have long shared the same underlying code to translate template data into the citations that readers see.  From  in February 2007 until  in April 2013,  used .  After the April 2013 edit,  has been using Module:Citation/CS1. Today,  and 16 CS1 templates use the Lua module. —Trappist the monk (talk) 19:40, 30 April 2014 (UTC)
 * Trappist the monk's comment is quite true. However, in the past, the Citation Style 1 templates issued many fewer error messages, and for those parameters that were checked, the checking was looser. Tightening of parameter meanings was discussed at Help talk:Citation Style 1, and documented. No such discussion or documentation occurred for the Citation template, except to include a statement that it now uses the Citation Style 1 Lua module. Jc3s5h (talk) 20:27, 30 April 2014 (UTC)
 * Error checking in the -rendered templates was limited to categorizing pages that used one or more of a list of deprecated parameters and to archiveurl and archivedate error messages. This error checking applied to both  and the CS1 templates.  More complete error checking came as part of the transition to Module:Citation/CS1 and applies to all templates that use it.—Trappist the monk (talk) 10:04, 1 May 2014 (UTC)
 * Information: A bit more historical data. Here are the dates when the big four of CS1 migrated first to  and then to Module:Citation/CS1.
 * core: ; module:
 * core: ; module:
 * core: ; module:
 * core: ; module:
 * —Trappist the monk (talk) 10:04, 1 May 2014 (UTC)
 * Comment: I believe that the whole point of citations is to help readers locate the sources being cited. Most of the red error messages are useful in all of these cite/citation templates because they alert editors to problems with the citation that make it harder for readers to locate the source being cited. For example, a large number of date errors in citations are there because the dates are ambiguous or erroneous, e.g. 3/10/12 or July 16, 2912. I do not think it would help readers if we stopped flagging and fixing such errors in citation.


 * I agree with the note above that "accessdate without URL" is a debatable error, but I don't see why it would be treated differently in citation from how it is treated in (e.g.) cite book. If we turn it off in one place, we'll probably turn it off everywhere. – Jonesey95 (talk) 05:36, 1 May 2014 (UTC)


 * I wonder if Jonesey95 has followed the citation style discussions over the years. Maybe I should review the results of those discussions. Wikipedia does not have any house citation style. Citation style should be consistent within an article, but a particular article could use Citation Style 1, Citation (which is a different style from CS1), APA, Chicago, MLA, IEEE, International Astronomical Union, or a style made up just for that article. The WP:Manual of Style does not control the date formats in citations; the particular citation style chosen for the article does. If one is using APA style, the publication date would be written "(2014, May 1)." If the Citation style is chosen for an article, there are hardly any date rules, because Template:Citation/doc contains hardly any date rules. The rules in Help:Citation Style 1 or Template: Cite book don't apply, because those are from a different style. Jc3s5h (talk) 05:50, 1 May 2014 (UTC)
 * That is helpful information. Thank you. I started working on citation errors about a year ago, so I do not have much historical knowledge of citation-related editing and discussion prior to that. I have no further comment at this time. – Jonesey95 (talk) 06:29, 1 May 2014 (UTC)
 * Thank you for opening this RfC and for providing the history. You mentioned "The WP:Manual of Style does not control the date formats in citations" - could you please help me understand this better?  Is this documented anywhere?  I don't see anything in Manual of Style/Dates and numbers saying that it does not apply to citations.  Thanks!  GoingBatty (talk) 23:33, 1 May 2014 (UTC)
 * In WP:MOS, have a look at these statements:
 * "* All the dates in a given article should have the same format (day-month or month-day). However, for citations, see Citing sources."


 * "* Citing sources – explains process and standards for citing references."


 * In WP:MOSNUM see this quote:
 * "* Publication dates in references should all use the same format. Any format from the "Acceptable date formats" table above may be used, unless the citation style being used requires a different format (however, all-numeric date formats other than must still be avoided)."


 * Jc3s5h (talk) 01:42, 2 May 2014 (UTC)


 * I got a notice to give comment on this issue but I do not understand the issue here. Could someone restate it in a different way? I feel that others might also be confused.  Blue Rasberry   (talk)  01:09, 22 May 2014 (UTC)
 * Agreed, Bluerasberry. I looked it over a while ago and couldn't make head or tails of it. Evensteven (talk) 01:13, 22 May 2014 (UTC)


 * It's understood that a citation made with {{tl}cite book}}, {{tl}cite journal}}, etc., will look different from a citation made with citation in at least one respect: the various elements will be separated by periods with the former and commas with the latter. Here are two examples, first done with {{tl}cite news}} and the second done with citation:




 * Notice, as an example, that the title of the article is separated from the title of the newspaper by a period in the first case but a comma in the second case. Everyone agrees about this difference.


 * But should other differences be allowed? It's been agreed that citation style 1 (cite news, cite book, etc.) shouldn't use any date format other than the ones listed as acceptable in the table in WP:MOSNUM. Does that apply to the citation template? Or would it be OK to give the publication date as "2013, October 17"? What about other differences? Jc3s5h (talk) 03:36, 22 May 2014 (UTC)


 * ,, does my attempt at clarification help? Do you have opinions on the issue? Jc3s5h (talk) 20:17, 28 May 2014 (UTC)


 * Disallow differences Thanks for the explanation, . I have previously documented policy proposals about citations and care about this issue. I want to encourage anyone to voluntarily do what they like to modify citations when they make a conscious choice to do so, but at the same time, I feel that people who just want to generate a citation and who have no opinion about the style or formatting of it should be directed to a single default, as is the case with the citation generated when autofilling from the RefToolbar or when using the most common templates. If someone wants to consciously deviate from the most common practices then they should be able to do so, but functionality to encourage that should not be part of the most common tools such as this template.  Blue Rasberry   (talk)  20:43, 28 May 2014 (UTC)


 * I do understand the clarification through to the points you list as having already been agreed (I think), and I agree with them also. As for "other differences", I'm no expert, or even very sophisticated, so my opinions are only for what they're worth (which is why I've offered so few until now). But: date formats agreed for CS1, but you're asking if it "applies" to citation? That I don't understand. Doesn't CS1 drive or implement citation? So what's to decide?


 * But my opinion on dates is stick with MOSNUM, and whatever form is used in the input, use the same form in the output. Date formats are an engineering rat's nest if not controlled, and I think it would be useless to try supporting them all anyway. If some editor(s) really want another format, let them argue for it to be included in MOSNUM. They're also a rat's nest for article text if not controlled. The last thing we want is for all the alternatives to have to be explained or differentiated in article text because "Americans do it this way and Europeans that way" or other time-wasters. No need for our programs to have to be slowed down parsing all that stuff either. I prefer "[d]d MMM yyyy" myself, but anything that is clear without explanation suffices.


 * As for any "other things", I am unfamiliar with the various citation "standards" that are out there, except as a reader who sees citations in a variety of materials. My impression is that they're something of a waste of time, because there are too many of them for any to be any kind of a real accepted standard. I also think that the whole point of citation is to make it clear how to go find the information source in a library or some other repository, at least as close as a page number. If a citation does that, I consider it successful, and don't bother about more. I consider myself to have a scholarly bent, but my practical side rebels against picky academicism and sophistry, both enemies of real scholarship in my opinion. On WP, my opinion is that we want to allow editors to use whatever prominent system they wish to choose, so if there are two or three (or more) clear ways citations are done "out there", then we ought to have support for it "in here". I have no opinion whether CS1 or citation or anything else is the right vehicle for "school A" or "school B". Only, don't close out options that are currently open and available somewhere. From an engineering standpoint, I'd say have CS1 try to emulate one prominent method, and have a CS2 for another method, up to CSn to cover all necessary methods. Does "CS2" currently exist under the pseudonym "citation" right now? I don't know, because I am completely confused about what CS1 is "for" and what citation is "for", and I'm frustrated that there is no documentation that answers such basic questions. So please try to parse what I say by realizing that I may not be using your terminology "correctly". I tend to prefer what citation currently does even for WP bibliographies, and I've seen bibliographies in any number of books that do them with comma separators, so it can't be that uncommon. Just don't eliminate preferences or methods overall, however things are implemented. Design with the idea of preserving the choices. Engineering engines tend to work best when they know if they're a race car or a Mack truck. But I'd suggest avoiding Winnebagos. I hope this helps a bit. It all seems very baseline and obvious to me, so I have a hard time imagining that it hasn't occurred to others already. Evensteven (talk) 21:10, 28 May 2014 (UTC)

, I think it's convenient for this discussion to think of four discussion domains:
 * 1) The whole world of citation, exemplified by your statement "On WP, my opinion is that we want to allow editors to use whatever prominent system they wish to choose, so if there are two or three (or more) clear ways citations are done "out there", then we ought to have support for it "in here"."
 * 2) The set of templates described at Help:Citation Style 1. This is a subset of the no. 1.
 * 3) The set of templates that ought to described at Help:Citation Style 2, but isn't. Since there is no acceptable description, either Help:Citation Style 2 should be fixed, or CS2 should be regarded as defunct, abandoned, and depricated. The documentation for the collection of templates that used to be called citation style 2 seems to have disappeared.
 * 4) The Citation template, documented at Template:Citation. Historically this has been a distinct style from CS1, although it is implemented with the same module.

As for the whole world of citations usable in Wikipedia, the "prominent system[s]" you describe contradict MOS and MOSNUM, so it is impossible to both allow all prominent systems and impose MOS and MOSNUM on citations in general. MOSNUM does apply to CS1, but only because the editors interested in CS1 have chosen to adopt that standard in Help:Citation Style 1.

I vehemently disagree with the concept that the development of a module should drive the appearance of a style. The editors who use a style determine what the style should look like, and the module coders write in accord with the consensus of the users of the style. If the users of the Citation style end up deciding they want the style to look so different from CS1 that the current Lua module can't accommodate the differences, it would be necessary to write a new module for Citation. Jc3s5h (talk) 21:45, 28 May 2014 (UTC) (Updated 0:10 29 May 2014 (UT).
 * uses only one template, and that is ; thus, your discussion domains nos. 3 & 4 are the same. -- Red rose64 (talk) 22:25, 28 May 2014 (UTC)
 * There were a group of templates with names of the form cite xxx which were somewhat different than the CS1 templates. But the documentation for that group seems to have disappeared. Help:Citation Style 2 describes the Citation template; I'm not sure there was consensus to apply that label to the Citation template. Jc3s5h (talk) 00:11, 29 May 2014 (UTC)
 * Well, as I said, take my comments for what they're worth, and it doesn't hurt my feelings if they're not worth anything. I wouldn't be surprised. I'm not asking for massive additions to support citation styles we don't already have, and I have no suggestions for expansion in the MOS or MOSNUM. I understand the what the discussion domain #1 above is (and did refer to it), and now I understand that domain #2 is a subset (I didn't know that). As for domain #3, I didn't know a CS2 actually existed when I wrote my comment above. As for domain #4, I wasn't aware that CS1 and citation have "distinct styles". Having seen that they are implemented with the same module, I assumed that they were of matched styles. But again, there's no documentation to clear up such confusions. So, thanks for your clarifying points, and sorry I wasn't able to say anything useful. Evensteven (talk) 00:36, 29 May 2014 (UTC)
 * The biggest difference between CS1 and CS2/citation is that CS1. Separates. Everything. By capital letters and periods. whereas CS2/citation, uses commas, between parts of the citation. A secondary difference is that CS2 automatically produces reflinks allowing the harv (parenthetical citation) templates to be used in conjunction with it. CS1 can also do this but only if you remember to include harv as part of the citation, and only if you're using one of the more major CS1 templates whose authors remembered to include that as one of the parameters it handles. —David Eppstein (talk) 01:04, 29 May 2014 (UTC)
 * How can that be right when CS1 implements citation, and citation's output looks more like CS2? CS1 must be able to do the commas as well as the periods, capitals, etc. Is CS1 designed to be a grand-daddy do-all-styles core? Evensteven (talk) 03:01, 29 May 2014 (UTC)
 * I'm going to ignore the part of your question about CS2, because in my opinion, attempting to change the meaning of CS2 from a collection of some of the templates with names like cite xxx to being a synonym for the Citation template has destroyed the term CS2.
 * Module:Citation/CS1 implements the Citation template and some of the CS1 templates. Some of the CS1 templates have not been converted to the module yet, and are still implemented with Template:Citation/core. It is possible to make the output of a CS1 template look like the output of a Citation template by adding the parameter " ".
 * Neither Module:Citation/CS1 nor Template:Citation/core are designed to be a grand-daddy do-all-styles core. Neither of them can duplicate any style from any printed style guide, such as the Modern Language Association, APA style, or the Chicago Manual of Style, but any of these could be chosen as the citation style for a Wikipedia article. Jc3s5h (talk) 03:31, 29 May 2014 (UTC)
 * When was ever "a collection of some of the templates with names like cite xxx"? When the page was, it only mentioned one template, ; subsequent edits have not altered that. -- Red rose64 (talk) 07:51, 29 May 2014 (UTC)


 * I have restored the text you removed from my comment. I recall that when Citation/core was introduced and many cite xxx templates were re-implemented through invocation of Citation/core, some of the remaining ones were referred to as CS2, because some of the remaining ones were quite similar to each other. But I am unable to find where that was written. Jc3s5h (talk) 11:44, 29 May 2014 (UTC)


 * I found what remains of Citation Style 2: Category:Citation Style 2 templates. Jc3s5h (talk) 12:09, 29 May 2014 (UTC)


 * While I appreciate the helpfulness and many pointers here, may I point out that this discussion itself underlines the usefulness of documentation? The drift over time in meaning of CS2, inconsistent among practitioners, illustrates how high-level definitions of terms and overview of concepts and directions is critical to the functioning of an engineering group itself, much less anyone who might use the software. It is difficult to design or discuss when confusions arise even among those most familiar with details and history. Consider then what it means to be unfamiliar. That's when you get demands from editors who want this or that item that they can't find. And without a baseline document for engineers to work from, it becomes difficult to see all the angles about how such a demand might fit into a whole. That document also stabilizes terminology, as well as concept. Think how much time that saves in the long run. (In the short run too, very often.) Evensteven (talk) 20:00, 29 May 2014 (UTC)

Use of quote = as plagiarism
I've assumed for years that quote = is intended for quoting verbatim the source cited for purposes of WP:Verifiability. Today my understanding was challenged in this edit. Would someone kindly take on the challenge of updating Template:Citation to avoid any future confusion. Thanks! -- DanielPenfield (talk) 23:39, 2 June 2014 (UTC)


 * I don't see plagiarism an issue, since the quotes are fairly short and the sources are acknowledged. I also don't see a copyright issue, since the quotes are fairly short and the quotation is highly unlikely to serve as an alternative to the purchase of a copy of any publication. However, it is not usually necessary to quote material unless it is too intricate or controversial to summarize. Another reason to quote a short passage is when an electronic text is being quoted and it lacks page numbers or other paraphernalia to locate the passage of interest; the reader can use the search facility of the viewing software to find the passage of interest. Jc3s5h (talk) 00:34, 3 June 2014 (UTC)
 * (edit conflict) The template documentation looks fine to me. I believe that was trying to say that the quotations were too long, exceeding the bounds of fair use. That's not plagiarism, which is representing someone else's work as your own; it's a potential copyright violation. – Jonesey95 (talk) 00:38, 3 June 2014 (UTC)
 * It's also still untrue. Using a sentence or two (as these quotes are) to give a context for a reference is well within the bounds of fair use. —David Eppstein (talk) 03:09, 3 June 2014 (UTC)
 * This depends on the size of the work you are quoting. Using a sentence or two out of a 500 page text book = Fair Use, using a sentence or two out of a 4 page children's book = copyright violation. ReformedArsenal (talk) 12:40, 3 June 2014 (UTC)
 * This question seems to have been pretty much resolved and explained over here, a miscue by a fatigued editor. Evensteven (talk) 18:39, 3 June 2014 (UTC)

Bug wrt listing for editors?
I have noticed that when neither chapter nor section parameters are present, the string "ed." follows a list of editors. However, when one of those parameters is present, then no such string follows the list of editors. Seems odd, and undesirable. Is this a bug? Evensteven (talk) 03:33, 26 May 2014 (UTC)
 * Example please. --  Gadget850talk 13:33, 29 May 2014 (UTC)


 * I'll try to provide examples.


 * First, a citation with just editors, no chapter:




 * Next, a citation with a chapter and chapter author


 * Jc3s5h (talk) 14:12, 29 May 2014 (UTC)
 * Using cite book, for comparison (same behavior):
 * – Jonesey95 (talk) 14:24, 29 May 2014 (UTC)
 * Just so I could study how the output looks under various conditions, I created "citations" with generic values for parameters, including a fairly wide list of parameters I thought I might use. However, I had to use valid urls and dois instead of generic values, in order to avoid errors. While I got the same sort of thing as in the examples above provided by Jc3s5h and Jonesey95, I later found that there is another factor here.
 * Here are examples with my full list of parameters represented, except in the third case, |chapter=CHAPTER (|section=SECTION) is removed (of course, |chapterurl= must also be removed). Note how they illustrate my incomplete diagnosis, as editors never have the "ed." string following:
 * Now, if we eliminate use of the |work=WORK parameter also, we reveal the mystery behavior:
 * Just for fun, here's "cite journal" instead of "citation", using "|journal=" (the synonym of "|work="):
 * — Preceding unsigned comment added by Evensteven (talk • contribs) 19:28, 29 May 2014‎
 * The template uses Module:citation/CS1 as does . That module arranges the output in different orders according to whether it thinks it's processing the data for a journal or a book. If you use  or  the module uses the classes passed through by the template to decide the output order; but if you use, the only thing that the module has to go on is the presence or absence of certain parameters. If journal (or work) was specified, it formats for a journal, i.e. using "in" and no "ed.", but if journal and work are not specified, it formats for a book, i.e. using "ed." and no "in" -- Red rose64 (talk) 19:45, 29 May 2014 (UTC)
 * Later observation: This may already be covered in other discussion, but may not have been made explicit. produces not only the same "output order" as, but the exact same text with regards to editors, namely no "ed." or "eds." if "|work=x" is specified. I can see how citation might parse its parameters to determine what type of source to format for, but cite book already knows, and still produces the same display. We should expect cite book to know the difference. Therefore, either it omits the editor identifiers on the basis of some norms of citation formatting, or else it is just mimicking the citation template and ignoring the information it has but citation lacks. (Because of "engineering consistency" perhaps?) It still seems like a weak reason for the behavior,  though. Evensteven (talk) 07:58, 5 June 2014 (UTC)
 * For books where each chapter or section has a different author, and there is an overall editor(s), we have . -- Red rose64 (talk) 11:03, 5 June 2014 (UTC)
 * This is as documented:
 * Just for fun, here's "cite journal" instead of "citation", using "|journal=" (the synonym of "|work="):
 * — Preceding unsigned comment added by Evensteven (talk • contribs) 19:28, 29 May 2014‎
 * The template uses Module:citation/CS1 as does . That module arranges the output in different orders according to whether it thinks it's processing the data for a journal or a book. If you use  or  the module uses the classes passed through by the template to decide the output order; but if you use, the only thing that the module has to go on is the presence or absence of certain parameters. If journal (or work) was specified, it formats for a journal, i.e. using "in" and no "ed.", but if journal and work are not specified, it formats for a book, i.e. using "ed." and no "in" -- Red rose64 (talk) 19:45, 29 May 2014 (UTC)
 * Later observation: This may already be covered in other discussion, but may not have been made explicit. produces not only the same "output order" as, but the exact same text with regards to editors, namely no "ed." or "eds." if "|work=x" is specified. I can see how citation might parse its parameters to determine what type of source to format for, but cite book already knows, and still produces the same display. We should expect cite book to know the difference. Therefore, either it omits the editor identifiers on the basis of some norms of citation formatting, or else it is just mimicking the citation template and ignoring the information it has but citation lacks. (Because of "engineering consistency" perhaps?) It still seems like a weak reason for the behavior,  though. Evensteven (talk) 07:58, 5 June 2014 (UTC)
 * For books where each chapter or section has a different author, and there is an overall editor(s), we have . -- Red rose64 (talk) 11:03, 5 June 2014 (UTC)
 * This is as documented:
 * — Preceding unsigned comment added by Evensteven (talk • contribs) 19:28, 29 May 2014‎
 * The template uses Module:citation/CS1 as does . That module arranges the output in different orders according to whether it thinks it's processing the data for a journal or a book. If you use  or  the module uses the classes passed through by the template to decide the output order; but if you use, the only thing that the module has to go on is the presence or absence of certain parameters. If journal (or work) was specified, it formats for a journal, i.e. using "in" and no "ed.", but if journal and work are not specified, it formats for a book, i.e. using "ed." and no "in" -- Red rose64 (talk) 19:45, 29 May 2014 (UTC)
 * Later observation: This may already be covered in other discussion, but may not have been made explicit. produces not only the same "output order" as, but the exact same text with regards to editors, namely no "ed." or "eds." if "|work=x" is specified. I can see how citation might parse its parameters to determine what type of source to format for, but cite book already knows, and still produces the same display. We should expect cite book to know the difference. Therefore, either it omits the editor identifiers on the basis of some norms of citation formatting, or else it is just mimicking the citation template and ignoring the information it has but citation lacks. (Because of "engineering consistency" perhaps?) It still seems like a weak reason for the behavior,  though. Evensteven (talk) 07:58, 5 June 2014 (UTC)
 * For books where each chapter or section has a different author, and there is an overall editor(s), we have . -- Red rose64 (talk) 11:03, 5 June 2014 (UTC)
 * This is as documented:


 * --  Gadget850talk 19:51, 29 May 2014 (UTC)


 * Thanks. No bug then. I know I read that documentation at some point, but that detail didn't sink in far enough and I didn't remember. Evensteven (talk) 00:02, 30 May 2014 (UTC)

The documentation still needs some work. It does not accurately describe this case yet (a book with author and editor but no specific chapter cited):

I'll give it some thought and see if I can give the docs a more complete treatment. – Jonesey95 (talk) 00:44, 30 May 2014 (UTC)
 * How is this example not compliant with the documentation? --  Gadget850talk 01:38, 30 May 2014 (UTC)
 * The documentation says:
 * If authors: Authors are first, followed by the editors and the included work, then the main work.
 * If no authors: Editors appear before the included work; a single editor is followed by "ed."; multiple editors are followed by "eds."; more than three editors will be followed by "et al., eds."
 * But this example has authors and a single editor is followed by "ed.", which is a combination of both descriptions. The examples above are essentially the same to my eye, but they do not show "ed." Here's the same citation with a chapter listed:




 * And here's the citation immediately above with no authors:




 * The documentation says "If authors" and "If no authors", but it does not explain that the existence of a populated chapter will determine whether "ed." is shown. Does that make sense? I might be reading something wrong.


 * Maybe it should say "If authors, and if a chapter/section/contribution is shown:" and "If no authors:". Does that make sense? As I said, I need to give it some thought. – Jonesey95 (talk) 02:02, 30 May 2014 (UTC)


 * Jonesey95 is also correct about the documentation in another way. It describes behavior in the absence of work= or journal= only, not in their presence. It's just that he didn't give an example of that, as I did above. Evensteven (talk) 02:08, 30 May 2014 (UTC)

I think a reasonable wiki-editor will expect that persons labelled as editors in the parameters will be labelled as editors in the rendered citation, and will read the documentation with that presumption in mind. Therefore there is a bug, since the documentation does not contain any explicit statement describing the circumstances under which editors will be treated as authors. Jc3s5h (talk) 12:20, 30 May 2014 (UTC)

Thinking about the problem further, this problem cannot be solved with the existing set of parameters. The existing set of parameters do not distinguish between the authors/editors of the included work (where "work" is ordinary English and not a parameter) and the authors/editors of the containing work. Thus the template attempts to guess which is which, but cannot do so reliably. Using the citation of the Explanatory Supplement above as an example, what if Seidelmann had been presented on the title page of the book as the book author, rather than the book editor? The template would have no way to deal with that. Also, whatever convoluted logic the template uses for its guesses will be hard to explain in the documentation.

My proposed solution: Depricate the use of templates for smaller works contained within larger works, and tell editors to type those citations by hand, because the installed base of existing template instances makes the problem unsolvable. Jc3s5h (talk) 12:36, 30 May 2014 (UTC)
 * I don't see what is wrong with the current behaviour. The editors are effectively labelled as editors by the presence of "in"; they can't be the authors if a contribution/chapter appears "in" a work attached to their names. I guess "ed." or "eds" could redundantly be added, but it would be redundant. In a citation like Jones, P. (1999), "Inventing the wheel", in Smith, R., History of transportation, Smith must be an editor with respect to Jones' contribution, regardless of whether Smith is an author of other parts of the book. The "in" makes that clear . Thus Smith, R. (1999), "Beyond the wheel", in Smith, R., History of transportation is different from Smith, R. (1999), History of transportation, "Beyond the wheel" (chap. 12). Peter coxhead (talk) 12:52, 30 May 2014 (UTC)


 * Typically, if one chapter in a book were contributed by a named author, all the other chapters were also contributed by named authors, and the people who put the book together are called "editors". But it doesn't have to be that way. It's possible for an author to write most of a book but have a few "guest chapters". Both Chicago Manual of Style (p 708) and APA style (p. 204) identify the editor of a multi-author book with the abbreviation "ed." (although neither gives an example of a book that has both a book author and a different chapter author). Jc3s5h (talk) 13:42, 30 May 2014 (UTC)
 * I agree that some style guides do add redundantly add "ed." or "eds" (indeed I give examples using it in my student referencing guide), but my main point remains that the current output from the template isn't wrong, merely in a non-redundant style. Peter coxhead (talk) 14:22, 30 May 2014 (UTC)
 * I do not see an issue here which justifies telling people not to use the templates. [There is the isue of not having the "(ed)" text 100% of the time, see below.]
 * If they are an author, they go in author/first/last. If they are an editor, they go in editor/editor-first/editor-last. If the person was both they go in both. If you are citing a chapter written by John in a book which was primarily written by Frank, but not that section, you cite it as John as the author and Frank as the editor (not an author for the chapter being cited).
 * How each person would be listed could be, and often should be, different for different sections/chapters of a book or if the entire book is being cited.
 * Maybe I missed it, but I do not see the issue which is being raised (beyond the lack of displaying "(ed)"). Either the person is an author for the specific material being cited, or not. Either they are an editor for what is being cited, or not. They can, of course, be both. If they have some other role, or if there is something that requires further explanation then others.&mdash; Makyen (talk) 18:51, 30 May 2014 (UTC); clarify 18:54, 30 May 2014 (UTC)

Design issue wrt listing for editors
Continued discussion, but new section heading, for convenience and clarity. We have moved on from a discussion of "bug" into a discussion of "design flaw". Personally, I accept Gadget850's contention that the software works as it was intended. Since the documentation did not properly describe that, the bug was in the documentation. But I also agree with Jc3s5h that there is at least potential for considering the current behavior to be a design flaw, for the reasons he states. And again, I agree with Peter that this issue can lead us into complex real-world situations that automated citation tools such as these templates will have great difficulty supporting to everyone's satisfaction. It's a reasonable point how to deal with "redundancy" in the citation. The biggest engineering problem is how to draw the line in how "redundancy" is defined, how to support it/avoid it, and if there is any one way to support the various treatments that the citers may wish to see. There comes a point at which automation just requires one to go along with what a program does, because the software becomes too complex to be reasonably maintainable or bug-free. So at some dividing line, I think Jc3s5h's approach of just having the individual editor create the citation by hand makes sense. The real question is: where is the dividing line? And there's one design issue.

One complicating factor is that some articles use a consistent style that depends on the "Harvard" anchors, citing in a Bibliography, then referencing via short footnotes (sfn or harvnb). Perhaps it wouldn't be bad to have a new(?) very simple template that takes a whole hand-written citation intact as text, but associates it with an anchor (|ref=), for use within such an environment.

But I tend to agree with Jc3s5h that editors are editors, and should be identified as such in all cases by the CS1 or citation templates. This is a relatively straightforward thing to implement in both engineering and documentation and should prove to be readily understandable by editors. The more a program tries to do for you, the more it hides things, and the less understandable and reasonable it tends to become as a result. Peter could use Jc3s5h's hand-written citation approach to address redundancy issues where they come up. So could anyone, to address any type of citation style issue.

I think that is a basic design philosophy that makes some sense, and one approach to implementing it here. I'm not expert in this at all, but I offer it as a relatively clean starting point. But with this (or a replacement philosophy), it should become easier to resolve what ought to be deprecatable in current features. Jc3s5h had one suggestion, but there might be others that are reasonable too. I guess I'm saying that "design" might not be the only thing to consider. A re-"vision" or "concept" adjustment might resolve these ideas just as well, in addition to providing opportunities for making the software more maintainable and describable. Evensteven (talk) 17:37, 30 May 2014 (UTC)
 * I agree that it would be better to have "(ed)" or "(eds)" always displayed rather than just "In". Using just "In" makes an assumption that the reader knows about the convention that to say "In" is indicating that the person/people listed following it are editors.  This is not a good assumption for us to make and probably leads to confusion for readers. It is not something the average reader would expect when also seeing citations on the same page which do say "(ed)". The small amount of redundancy of also showing (ed)/(eds) improves communication and costs very little for those who are aware of the convention. I recall that when I first saw it, I did not immediately realize that was what the text was intended to communicate. While it was easy to figure out, it was not clear at first glance. In general, it should be clear at first glance for readers generally familiar with citations without the need to know some special convention. &mdash; Makyen (talk) 18:51, 30 May 2014 (UTC)
 * I don't see a reason why the harv references would be negatively impacted by properly citing editors as editors. It may make formatting the citation slightly more complicated as you may need to add a ref= section, but that isn't that big of a deal. ReformedArsenal (talk) 18:59, 30 May 2014 (UTC)
 * You mention "a new(?) very simple template that takes a whole hand-written citation intact as text, but associates it with an anchor (|ref=)" - we already have this, it is, see my comments of 16:09, 21 May 2014 at User talk:Magioladitis, the bit beginning "It's not necessary to use if you want to use ". -- Red rose64 (talk) 20:34, 30 May 2014 (UTC)
 * Thanks, all. I just keep on learning. Makyen, count me as one who didn't realize that implication of "In". I would guess that many (if not all) of the major citation styles grew within the academic community, but my fields of study did not require much more than a basic understanding of citation; they required more field application, less writing of papers. So in that regard, I probably have more in common with the average WP reader when it comes to parsing an existing citation. On that basis, I agree with Makyen that "(ed)/(eds) improves communication and costs very little" and with ReformedArsenal that there's no significant impact on harv references that I can see. Redrose64, thanks for the pointer to . Just what I was envisioning. Cheers. Evensteven (talk) 21:43, 30 May 2014 (UTC)

Publishers, editors, authors, etc. should allow hyperlinks
Authors, editors, publishers, etc. should have the ability to have hyperlinks on external websites. If I am citing another website, and I want to reference the author's page on that website, I should be able to do that. If I am citing an external website, I should be able to cite the "About Us" page for the publisher. Even if I use the correct wikitext for an external website, the citation gets messed up. — Preceding unsigned comment added by Zzmonty (talk • contribs) 02:19, 8 June 2014 (UTC)
 * I don't think that's a good idea. For one thing it doesn't conform to WP:ELNO: The external page for an author or publisher is unlikely to contribute useful information on the specific topic of an article. All of these fields do allow wikilinks to Wikipedia articles about the author or publisher, which should be sufficient for most purposes. —David Eppstein (talk) 04:07, 8 June 2014 (UTC)
 * I agree with . This has the potential to turn citations into spamfests. Who is going to ensure that all such links are germane and not an abuse of policy? Philg88 ♦talk 04:59, 8 June 2014 (UTC)
 * Likewise. If an author (etc.) is notable enough to have a WP page then use the existing link parameter. Freely linking to external websites also gets into WP:RS issues. I wonder if perhaps the OP has any particular idea of what purpose would be served by such external links. ~ J. Johnson (JJ) (talk) 21:49, 8 June 2014 (UTC)

Date range produces an error
If the date is a range:


 * date      = 2001 - Present

the error "Category: CS1 errors: dates" is produced. For anything on the web, a date range is valid, because the text is fluid and changes over time. Look at the history of any Wikipedia article. Although the access date is snapshot of the webpage, the actual website is a range of dates. — Preceding unsigned comment added by Zzmonty (talk • contribs) 02:15, 8 June 2014 (UTC)


 * The point of the citation is to identify when the website said whatever the editor saw when the editor read it. It would not be appropriate to indicate any date later than the date when the editor was reading it, because we have no idea if the website might be modified after the editor read it so that it no longer supports the statement in the article. Jc3s5h (talk) 23:38, 8 June 2014 (UTC)

Citing foreign-language patents
While working on IMSI-catcher, I noticed a German-language patent present in the references that was untagged with the language. But when I went to add the language and trans_title parameters, I found that they do not appear to work. There's no mention of language for citing patents, so is there something causing these fields not to work? — Sasuke Sarutobi (talk) 11:51, 22 July 2014 (UTC)
 * For patents, the recognised parameters are different from those used when citing books, journals etc., and there are a lot fewer as well. There are a total of 28 recognised parameters, 15 of which have up to 6 defined aliases. But none of these are language or trans-title. -- Red rose64 (talk) 16:18, 22 July 2014 (UTC)

Citing a US Plant Patent
Anyone know how to use the Citation template with a US Plant Patent? For example, this one. All the patent-number values I've tried produce an incorrect link to espacenet.com. Peter coxhead (talk) 06:53, 11 August 2014 (UTC)
 * Sorry to ask an obvious question, but have you tried to use cite patent? It looks like US patent reference may also be useful, though I have never tried it. (See also the See also sections in those templates' documentation.)– Jonesey95 (talk) 01:31, 12 August 2014 (UTC)
 * Yes, I've tried it and several other templates. In all cases the link produced is wrong. The patent number which works in Espacenet (but with different other parts to the URL than the one generated by the templates) is 13548P3. If you check the links at the following you'll see the issue:
 * Correct link http://worldwide.espacenet.com/publicationDetails/originalDocument?FT=D&CC=US&NR=13548P3
 * produces
 * produces (you need inventor to use citation
 * I've tried "13548P3", "PP13548", "PP13548P3" (and probably some others I've forgotten); none produce the correct link. Peter coxhead (talk) 07:59, 12 August 2014 (UTC)
 * It looks like works if you change part of the resulting URL from   to  . I don't know how to implement that in the template itself. I imagine that you could set a   variable that would then modify the URL appropriately, but I don't know the URL variable scheme on the Espace side to know what makes   different from  . – Jonesey95 (talk) 18:22, 12 August 2014 (UTC)

Bringing separator use between jstor, mr, etc., in line with Template:Cite journal?
Cf. Weil–Petersson_metric, which uses Citation, and the Cite journals at Maryam_Mirzakhani. Re. parameters like doi, mr, zbl, etc., Cite journal displays: "doi: φ. MR ψ. Zbl η." But Citation displays: "doi: φ, MR ψ, Zbl η" (comma separators; no final punctuation mark). I would suggest brining Citation use in line with Cite journal.  It Is Me Here  t /  c  09:56, 15 August 2014 (UTC)
 * Cite journal is part of . CS1 templates are a different style from citation and the two sets should not be mixed in an article. --  Gadget850talk 10:36, 15 August 2014 (UTC)
 * Right, but there's a huge gap between "should not" and "are not". Is there any value to preserving the extra commas seen with these identifiers in Citation? Do we have large numbers of articles that intermix Citation with rigorously correct hand-coded citations that use these identifiers? It seems like we're discussing a real edge case. LeadSongDog come howl!  15:42, 15 August 2014 (UTC)
 * My understanding was that CS1 templates are for editors who prefer full stops as separators in citations, and citation for those who prefer commas. Neither style is superior to the other, and either may be used in articles so long as in a particular article only one style is used. (Personally, I prefer citation because I feel that using full stops as separators makes it difficult to distinguish between citations if there are more than one in a footnote. But that's just me.) — SMUconlaw (talk) 15:59, 15 August 2014 (UTC)
 * has always used commas to separate the various parameters in the rendered citation. Similarly,  templates, like, have always used periods to separate the various parameters.   has always been rendered without terminal punctuation while CS1 templates default to a terminal period.  In both cases, should editors choose to do so, these default conditions can be overridden with the parameters separator and postscript.  See the template documentation for details.
 * I think that what Editor Gadget850 was describing is the mixed use of templates with  or other CS1 templates within the same article.  Similarly, templated references should not be mixed with non-templated references per WP:CITEVAR.
 * —Trappist the monk (talk) 16:06, 15 August 2014 (UTC)
 * Yes, I believe I understood what Gadget850 was saying, that they "should not" be mixed, but in point of fact they often "are" mixed. All I am suggesting is that we can minimize the problem that mixing causes if we simply eliminate the apparently valueless punctuation inserted between these identifiers. It is not just a cosmetic issue, it has real (if infrequent) implications in that an added period or comma at the end of an identifier can be misconstrued as being part of that identifier, potentially causing a search to fail. If the citation template rendered this detail in the same way as the CS1 templates, what harm would be done? LeadSongDog come howl!  16:39, 15 August 2014 (UTC)
 * Can you show us a concrete example where the inter-identifier punctuation causes problems?
 * I'm confused. At first you seem to be saying that all inter-identifier punctuation should be removed (eliminate the apparently valueless punctuation inserted between these identifiers). But, you then write: If the citation template rendered this detail in the same way as the CS1 templates,... which seems to argue for the CS1 style which, by default, separates the various identifiers with a period followed by a space.  Can you clarify me please?
 * —Trappist the monk (talk) 16:57, 15 August 2014 (UTC)
 * Yes, LeadSongDog, I'm also not clear what you are suggesting. I don't think you are arguing that punctuation between the identifiers should be eliminated entirely (and replaced by spaces?) even though you call them "apparently valueless", because you seem to be suggesting that they be separated by full stops. Actually, separating them using commas would arguably be better since DOIs contain full stops in them but not commas! But I'm not yet seeing a strong reason for departing from the current convention, which is full stops for CS1 templates and commas for citation templates. — SMUconlaw (talk) 18:08, 15 August 2014 (UTC)
 * If all the refs in the article use the templates, new citations should also be.
 * If all the refs in the article use, new citations should also be.
 * If the article has a mixture of and, go with the majority. Ideally, you should switch the minority form to match the majority: however, this cannot always be done. Switching from one to the other is usually a simple case of changing the template name -   to   (or vice versa) - because all of the parameters used by common  templates like , , and  are also recognised by . However, this is not always the case for some of the more obscure  templates. -- Red rose64 (talk) 19:27, 15 August 2014 (UTC)

An example may be in order. vice
 * Please note that both use a link icon after the redundantly bluelinked identifier (and url). Citation follows the linkout icon with a comma while Cite journal follows it with a period. In neither case is the punctuation useful. A space would suffice. LeadSongDog come howl!  20:05, 15 August 2014 (UTC)
 * Why are the links redundant? -- Red rose64 (talk) 20:13, 15 August 2014 (UTC)
 * Why are the links redundant? -- Red rose64 (talk) 20:13, 15 August 2014 (UTC)

Some or all of these previous discussions may be relevant: Headbomb {talk / contribs / physics / books} 21:21, 15 August 2014 (UTC)
 * Template_talk:Citation/Archive_5
 * Template_talk:Citation/Archive_4
 * Template_talk:Citation/Archive_4
 * Then if one or more of the ISSN, PMID, DOI links are redundant, don't include that parameter. None of them are mandatory. -- Red rose64 (talk) 22:01, 15 August 2014 (UTC)
 * I think that when a paper has a PMID or a DOI it should be included in the citation (and when both exist both should be included). ISSN is much less useful, though, because it identifies only the journal and not the individual paper within it. —David Eppstein (talk) 22:34, 15 August 2014 (UTC)

To clarify my reply to the original question, WP:CITEVAR states that you should use a consistent citation style within an article and that articles may be edited for a consistent style. As to why citation and CS1 are different, that is lost in template antiquity. --  Gadget850talk 00:32, 16 August 2014 (UTC)
 * It isn't that the links are redundant. It's that the link out icon (the square with an angled arrowhead to the right of the blue linked, underlined identifier) which is redundant. Either would suffice. If an editor wants to furnish ten identifiers for different ways to find the source, great. Why do we turn that into twenty things to click on? Do we think readers can only recognize blue links outside the reference list?
 * Yes, CITEVAR suggests a few such things, but if you think that that guideline has been widely implemented, I suggest clicking Random article a few times and seeing what pops up. One is far more likely to find an article which is completely unreferenced than one that has consistent references. Naked urls are still rampant. Hand coded refs mingle some with unlinked titles and others with the link spanning the entire ref. Sure there are a few FA or GA-class exceptions, but the mass of stubs, Start-class and C-class articles are a different story. In comparison, the differences between citatation and cite journal are minuscule, but shrinking the differences is still worth doing. LeadSongDog come howl!  04:11, 16 August 2014 (UTC)
 * The external link icon is not a function of CS1. Changing how external links are rendered in article text or in references is, I think a topic for another forum.  I think that the external link icons are important because they help readers to more easily distinguish between internal and external links.
 * —Trappist the monk (talk) 10:13, 16 August 2014 (UTC)
 * What Trappist said. Also, external links normally appear with the boxed arrow icon (which is similar to External.svg drawn half that size, although not identical), whether they are in references, the "External links" section, or inline in the article text - it's a consequence of the MediaWiki software adding the  class to those links. It is possible to suppress the boxed arrow icon, by means of the   class (or the  template), but this is an uncommon technique. In Today's Featured Article, for example, the Club home page link in the infobox has the boxed arrow icon, as do several dozen in the references - but of particular interest are the eight bullets under "External links". These eight bullets are followed by a total of thirteen links, of which two are normal internal links; one is an interwiki link (to commons:Category:Aston Villa FC) - these count as internal; six are normal external links which bear the boxed arrow icon; and four are external links where the boxed arrow icon has been suppressed by the use of the   class within the  template. -- Red rose64 (talk) 13:44, 16 August 2014 (UTC)
 * And the link rendered by url has an external link icon. You can suppress the icons per Help:External link icons. --  Gadget850talk 14:46, 16 August 2014 (UTC)
 * So the point comes down to this: neither the comma nor the period is needed between the external link icon for one identifier and the following identifier name. They just add to the visual clutter. We should lose them both from the citation and the cs1 templates.LeadSongDog come howl!  06:41, 17 August 2014 (UTC)
 * I think the external link icon is important for identifiers; without it links created by e.g. title url would be presented differently. Given that authors, journal titles, etc. are quite often wikilinked, it is useful to distinguish between internal and external links in citations. I guess the full stop or comma between identifiers could be suppressed, but it seems to me a very minor matter (and anyway it should be present in the printed version where the URL is given in full). Peter coxhead (talk) 08:27, 17 August 2014 (UTC)
 * Why should it be present in the print version? Does it serve the reader in some way? — Preceding unsigned comment added by LeadSongDog (talk • contribs) 19:13, 17 August 2014‎
 * Yes, if it's been printed and given to somebody else who then wants to follow the link - they get to a computer, read the URL off the page and type it in. -- Red rose64 (talk) 07:50, 20 August 2014 (UTC)
 * That's a reason to have the URL on the page, not a reason to have the icon AND the comma/period on the page. Are you suggesting that users are more likely to misread the URL without those features? LeadSongDog come howl!  14:50, 20 August 2014 (UTC)
 * What exactly is your objection? You would rather have instead of, ,? Because you do need to separate identifiers from each other, just like you would seperator authors (B Smith J Jones E Elbert vs B Smith, J Jones, E Elbert). Headbomb {talk / contribs / physics / books} 15:53, 20 August 2014 (UTC)
 * The boxed arrow icon doesn't appear on printed pages. -- Red rose64 (talk) 16:18, 20 August 2014 (UTC)

I would certainly rather have no comma than the mix of comma-alone, boxed-arrow-plus-comma, boxed-arrow-alone, boxed-arrow-plus-period, and period-alone that presently arises when citation types are mixed (as they usually are, despite the guideline). On printed pages, the URLs are underlined, with a break in that underlining serving to visually separate adjoining URLs. No added boxed-arrow icon nor punctuation are needed to cue the reader that xhe is looking at distinct strings. However, since we seem to be going around in circles here, and no other editor seems to see this as a concern, it seems like time for me to drop the stick and leave the leavings time to ferment.LeadSongDog come howl!  19:31, 20 August 2014 (UTC)

Ending punctuation
How do you suppress the period at the end for usage in a string cite? postdlf (talk) 21:35, 19 September 2014 (UTC)
 * By default, the citation template does not produce a period at the end. Are you perhaps referring instead to the Help:CS1 templates such as cite journal? If so, the answer is to use none. But because their formatting is a little different, those templates shouldn't be mixed with citation. —David Eppstein (talk) 22:38, 19 September 2014 (UTC)

Google books
Editors could save much tricky syntax with templates like:

and

In my dreams the former would work like, producing a global, editable reference, and the latter would link both to the cite and separately (using the page number) to the appropriate page number in Google books.

Please let me know if there is another place to put this suggestion. Thanks. Lfstevens (talk) 21:47, 20 September 2014 (UTC)


 * cite doi relies on a bot to fill in the citation. --  Gadget850talk 23:13, 20 September 2014 (UTC)
 * A bot which (as I understand it) went down with Toolserver at the end of June. But that aside, please can we avoid creating yet more citation templates? See small-font comment by PBS dated 19:06, 20 September 2014 at Wikipedia talk:Citing sources -- Red rose64 (talk) 23:30, 20 September 2014 (UTC)


 * Made some comments over there. I don't understand the reticence towards enhancing the best-working part of the ref system&mdash;cite.

Accessibility and COinS
There may be an accessibility issue with the COinS metadata that is appended to citations emitted by the template as well as all the  templates. See Village pump (technical)/Archive 130. -- Red rose64 (talk) 22:58, 24 September 2014 (UTC)

Access date. Again
Where was discussed? The bit in parenthesis was removed and re-added today, and it appears to be the crux of a small edit-war at St Pancras railway station. -- Red rose64 (talk) 20:49, 23 September 2014 (UTC)
 * I do not recall that text being discussed anywhere. It doesn't make sense to me, and it is inconsistent with the more verbose documentation for accessdate that appears farther down:
 * "Full date when original URL was accessed; use the same format as other access and archive dates in the citations; requires url.[date 1] Do not wikilink. Not required for web pages or linked documents that do not change; mainly for use of web pages that change frequently or have no publication date. Can be hidden or styled by registered editors."
 * I would recommend removing that parenthetical remark unless there is some strong consensus on the Talk page for its insertion. – Jonesey95 (talk) 03:04, 24 September 2014 (UTC)
 * Agreed, and I have done so. (1) In some cases the identifiable date is explicitly a starting point (e.g. "2001 onwards", so even in its own terms the addition isn't correct. (2) There's no way of guaranteeing that user-maintained web sites weren't updated after any stated date – it's easy to forget to change the date. (3) There's also the issue of link-rot; an access date is a useful indicator.
 * Actually I think the documentation for accessdate is too weak in this respect. I would prefer to see an access date unless there is a clear 'permanent' link such as a doi, pmid, etc. Peter coxhead (talk) 06:49, 24 September 2014 (UTC)
 * The accessdate gives context about when that section of article was contributed/revisioned; external websites do change, which is why we have sites like www.newsdiffs.org; accessdate is helpful when rescuing and selecting the nearest archiveurl. In the age of the web and most sites being in content management systems it is not enough to re-state an apparent publication date.  The advice we give to people wanting to cite Wikipedia is to always use a stable URL with oldid to a specific revision (ie. the accessdate + time).
 * Since accessdate is (nearly) always known by the contributing editor, the status-quo with near-universal addition of accessdate should be codified more strongly than date which may be known. We have had a situation where the present/temporary IP wording has been used as the basis of for widescale removal of accessdate from articles; something I would prefer to see unambiguously prevented, if nothing else that simply to prevent needless unhelpful diff-noise and the continued time-wastage from trying to work out what really changed.  Yes, I would support a stronger statement. —Sladen (talk) 07:58, 24 September 2014 (UTC)
 * While accessdate may not always be relevant, removing it is certainly never helpful. My personal opinion is that any web links should include accessdate, as even stable websites change or disappear eventually. -mattbuck (Talk) 22:21, 24 September 2014 (UTC)
 * No, if it's not relevant it's helpful to remove it, because it creates needless clutter. Nikkimaria (talk) 04:17, 25 September 2014 (UTC)
 * "Clutter" – possibly, although it's only a small part of a good citation; "needless" – no, as has been explained above. Peter coxhead (talk) 09:36, 25 September 2014 (UTC)
 * Yes, if it's truly a document that does not change. Nikkimaria (talk) 14:24, 25 September 2014 (UTC)
 * And will not change its location. Basically this means documents with DOIs or DOI-like URLs. Peter coxhead (talk) 14:51, 25 September 2014 (UTC)
 * Even if it is clutter, having an accessdate is doing no harm. For this reason, I think they should not be removed from citations. Mjroots (talk) 20:09, 25 September 2014 (UTC)
 * Clutter is arguably harmful, and if it's doing no good there should be no problem with removing it. Nikkimaria (talk) 20:34, 25 September 2014 (UTC)
 * accessdate is highly useful and contextual, both for our editors and for other editors who contribute to an article later in (ie. "doing good"), and so—if I understand your argument correctly—should not be removed. —Sladen (talk) 01:26, 28 September 2014 (UTC)
 * I agree with Nikkimaria. I have never found accessdate to be of the smallest use. -- Alarics (talk) 12:41, 28 September 2014 (UTC)
 * It is useful when fixing deadlinks, as it is relevant to finding the correct archived snapshot per wp:SAYWHEREYOUGOTIT. Also, it comes in handy when it is necessary to decipher which rev of the article the citation was used to support. Editors often change the meaning of assertions (intentionally or not) while leaving the existing citation in place... LeadSongDog come howl!  05:13, 29 September 2014 (UTC)
 * And there are editors who believe you need access dates for hard copy sources. And those who will go through an article, check to make the links work and update the access dates. Regardless, this is a perennial discussions and as usual will not end with any consensus. I hide the access dates and don't worry about them. --  Gadget850talk 13:08, 29 September 2014 (UTC)

Where to put archive database used to access article?
I put in citations in the publisher field: "Accessed via LexisNexis" or "Accessed via NewsBank", as the case may be, to denote where I found the source and where others in the future may access the source if they wish to read that article text.

Is there another place or another field to place this info?

I ask because sometimes people with bots or scripts come by and remove the entire "publisher" field, and then my notes on how to access those articles would be lost !

Any help?

&mdash; Cirt (talk) 00:18, 1 October 2014 (UTC)


 * Use via like this:


 * Because what you put in publisher is included in the citation's COinS metadata, using it to hold information that has nothing to do with the actual publisher corrupts that data. via is not part of the COinS.


 * —Trappist the monk (talk) 00:41, 1 October 2014 (UTC)
 * Oh, wow, thanks very much!!! &mdash; Cirt (talk) 00:44, 1 October 2014 (UTC)
 * Now I just gotta find all the places I used "publisher" instead of "via", and fix those before the bots and/or scripts take away those publisher fields!!! &mdash; Cirt (talk) 00:48, 1 October 2014 (UTC)

Might help if we documented via. Anyone have a good description of use? --  Gadget850talk 01:07, 1 October 2014 (UTC)
 * Yes that would be most helpful! And if it could be publicized to the people who run bots and scripts not to remove it? &mdash; Cirt (talk) 01:55, 1 October 2014 (UTC)
 * My mistake: this is documented. --  Gadget850talk 11:42, 1 October 2014 (UTC)

, can someone please help fix q:Template:Cite news and q:Template:Citation at Wikiquote so they display the via parameter in cites? Thank you so much, &mdash; Cirt (talk) 20:18, 9 October 2014 (UTC)
 * Not that simple. Those templates still use citation/core while we have switched to Module:Citation. This would need the module and sub-modules copied over, then the templates can be updated. --  Gadget850talk 20:40, 9 October 2014 (UTC)
 * Can you help with that please, ? &mdash; Cirt (talk) 20:42, 9 October 2014 (UTC)

Two user script questions
Maybe someone here can answer these questions there Wikipedia_talk:User_scripts. &mdash; Cirt (talk) 19:40, 10 October 2014 (UTC)

"produces output identical to that of the Cite templates [..] Cite web" - no
See my edit. I noticed ", retrieved" in lower case. It was because of "citation" and changed to upper case (and period) when I changed to "cite web". I didn't change all the other cases as I guess it can be "fixed" here (I couldn't figure it out). I'm not even sure if it's preferred to change in pages. comp.arch (talk) 18:05, 11 October 2014 (UTC)
 * and citation are two different styles, so you should use the current prevailing style. The article is now a mix of, citation and non-template citations, thus the rendered styles are going to differ. --  Gadget850talk 18:28, 11 October 2014 (UTC)


 * uses a comma separator so it is considered, as I understand it, to be a sentence fragment. As such, 'retrieved ...' is not capitalized.  The CS1 citation,, etc use period (full stop) separators so 'Retrieved ...' Starts its own very short sentence.  There are those who prefer one way; there are those who prefer the other.  Whenever one style is used, the other should not be – use should be consistent throughout the article.


 * It is possible to make CS1 mimic by using, and none:
 * and the other way as well:
 * – here . and.
 * and the other way as well:
 * – here . and.
 * – here . and.


 * Why? I don't know.


 * —Trappist the monk (talk) 18:30, 11 October 2014 (UTC)

PD/WS icons
Hello. I was wondering if it would be possible to add a parameter to set an icon (like Template:Cite_DNB) for works that are in PD and/or in wikisource. That way editors can easily spot "free" sources to improve the page. Of course I'm not talking about works that are already incorporated in the article (i.e. Template:Source-attribution). Thanks. -Ihaveacatonmydesk (talk) 22:01, 18 October 2014 (UTC)
 * Is this a standard internationally recognized icon or something we made up?
 * The icon is currently not accessible; that is, it has no meaning to our readers who use screen readers. --  Gadget850talk 09:55, 19 October 2014 (UTC)


 * Can we? Yes. Should we? No. Adding icons to an already dense collection of text, numbers and punctuation, is just unnecessary clutter.


 * Unlike the pdf icon that is added by css, the icons at templates like can easily and so should have alt text added to them.


 * —Trappist the monk (talk) 10:40, 19 October 2014 (UTC)
 * Anthropomorphism is an example of something like that already happening. It doesn't seem cluttered to me. - Ihaveacatonmydesk (talk) 12:00, 19 October 2014 (UTC)
 * I see a blue ladybug on ref 7 and a cents symbol on 8. --  Gadget850talk 12:04, 19 October 2014 (UTC)
 * That seems more like an issue with the way those templates deal with icons - Ihaveacatonmydesk (talk) 12:27, 19 October 2014 (UTC)
 * Anyway, I see your point and I'm all for alt text of course. - Ihaveacatonmydesk (talk) 12:32, 19 October 2014 (UTC)

Formatting of journal article title
Has the formatting of title in journal articles formatted by citation changed recently or has it always been broken? It should be in an upright font, in quotation marks, but instead it comes out italicized (like the title of a book). Example: There are a lot of examples of citation of journal articles that use title, and our documentation says to use title in this type of citation, so I'm pretty sure this is indeed a bug and not just a bad choice of input parameters. —David Eppstein (talk) 02:54, 19 October 2014 (UTC)
 * Cite journal, formatted correctly (except for the separator being a period):
 * CItation with title, formatted incorrectly (title in italics rather than quoted):
 * CItation with contribution, formatted correctly:
 * CItation with title, formatted incorrectly (title in italics rather than quoted):
 * CItation with contribution, formatted correctly:
 * CItation with contribution, formatted correctly:
 * CItation with contribution, formatted correctly:
 * It does seem to have changed, yes. This might be the same issue as Help talk:Citation Style 1. -- Red rose64 (talk) 08:38, 19 October 2014 (UTC)


 * I think this also affected trans_chapter. For journal articles, one would set title and use trans_chapter but now one must use trans_title instead. Many editors have done clean up, so now I think all the trans_chapter params have been changed. I don't know if that was the intended result. Glrx (talk) 20:29, 22 October 2014 (UTC)


 * I'm confused. The correct use of trans-title has always been with title; trans-chapter with chapter.  The formatting issue described by Editor David Eppstein with regard to  has been fixed in the sandbox.
 * Making of changes based solely on current live module's improper formatting of title parameters is strongly discouraged.
 * Making of changes based solely on current live module's improper formatting of title parameters is strongly discouraged.


 * —Trappist the monk (talk) 21:46, 22 October 2014 (UTC)


 * Template talk:Citation/Archive 6 Glrx (talk) 02:13, 23 October 2014 (UTC)


 * Now I'm even more confused. Why are you bringing up an old conversation from when citations used ?  I have lifted a citation from that conversation that was described as not working properly.  Here is a side by side comparison of the old, the live Module:Citation/CS1 and new Module:Citation/CS1/sandbox:


 * —Trappist the monk (talk) 03:04, 23 October 2014 (UTC)

Is anything happening with moving this from the sandbox to the live template? —David Eppstein (talk) 20:59, 12 November 2014 (UTC)


 * Because Module:Citation/CS1 is used in almost 2.5 million pages I try not to make multiple incremental changes to the live module. Unless the problem is a showstopper (the module uses up all of its allotted time, there is a fatal error, etc) incremental changes are made to the sandbox. After enough changes have accumulated, then I will update live module. Because there are several changes in progress, the sandbox in not ready to do an update.


 * This has not been forgotten.


 * —Trappist the monk (talk) 22:25, 12 November 2014 (UTC)
 * Ok, thanks. I'll just continue to wait patiently then. —David Eppstein (talk) 23:56, 12 November 2014 (UTC)

Issues from 29 November 2014 update
Good to see the formatting of journal articles using the Citation template fixed!

A couple of issues which I think are in response to the latest update.
 * The example of a conference paper in the Citation template documentation needed to be changed. It was:
 * which used to use url as the link for contribution but now uses it as the link for title. I've updated the example to the now correct contribution-url but any citation using the previously recommended parameters will now be slightly wrong, with no error message.
 * which used to use url as the link for contribution but now uses it as the link for title. I've updated the example to the now correct contribution-url but any citation using the previously recommended parameters will now be slightly wrong, with no error message.


 * Using cite web it's possible to get a citation with just the title in double-quotes:
 * There seems to be no way to achieve this using the Citation template; you have to have website to get this formatting of the web page title:
 * Ideally changing "citation" to the appropriate "cite XXX" would produce the same output.
 * Ideally changing "citation" to the appropriate "cite XXX" would produce the same output.
 * Ideally changing "citation" to the appropriate "cite XXX" would produce the same output.
 * Ideally changing "citation" to the appropriate "cite XXX" would produce the same output.

Peter coxhead (talk) 17:57, 29 November 2014 (UTC)


 * Yep, saw your corrections to the citation template documentation. contribution-url (or any of the chapter-url aliases) is the correct fix.  The 29 November update consolidates all of the chapter parameters.


 * I'm pretty sure that without work didn't quote the title before the 29 November update.  There isn't enough information in a title and url for Module:Citation/CS1 to know because that same cite could be a book title with link to a scanned copy at archive.org.


 * Here is a comparison between the old version of  and the current Module-based version


 * —Trappist the monk (talk) 19:10, 29 November 2014 (UTC)


 * Ah, but in the past you could use contribution to produce the effect that you can with as per this comparison:
 * My point is that there should be always be way to produce CS2 versions of CS1 styles if CS2 users aren't to become even more of second-class Wikipedia citizens than they are now. One answer would be to allow none, which provides the required information. Peter coxhead (talk) 20:03, 29 November 2014 (UTC)
 * My point is that there should be always be way to produce CS2 versions of CS1 styles if CS2 users aren't to become even more of second-class Wikipedia citizens than they are now. One answer would be to allow none, which provides the required information. Peter coxhead (talk) 20:03, 29 November 2014 (UTC)


 * I'm not sure I see a difference between none and Some Website (it is the name, not the domain name: Wikipedia not Wikipedia.org). Adding the website parameter is something that users of both CS1 and CS2 should be doing.  CS2 has the same 'problem' with journal and news cites when journal or work is empty or omitted}}:
 * When making these types of cites, editors should include journal or other work alias regardless of whether they use CS1 or CS2.
 * When making these types of cites, editors should include journal or other work alias regardless of whether they use CS1 or CS2.
 * When making these types of cites, editors should include journal or other work alias regardless of whether they use CS1 or CS2.


 * —Trappist the monk (talk) 20:19, 29 November 2014 (UTC)


 * There's a difference between websites and journals or newspapers. The latter always have a sensible value for the "work" parameter. Websites certainly don't always have a "name" for the website as a whole. I see that some misguided editor used a web page I maintain as a reference at Sutton Coldfield, with the citation:
 * which (supposing it counted as a reliable source) looks a reasonable citation. How can this be set up using ? What would you use for website? Articles about places or popular culture are full of citations to websites which don't have clear overall titles. Peter coxhead (talk) 21:34, 29 November 2014 (UTC)
 * which (supposing it counted as a reliable source) looks a reasonable citation. How can this be set up using ? What would you use for website? Articles about places or popular culture are full of citations to websites which don't have clear overall titles. Peter coxhead (talk) 21:34, 29 November 2014 (UTC)


 * Where did the section edit link go for this topic?


 * Like this?
 * —Trappist the monk (talk) 21:54, 29 November 2014 (UTC)
 * —Trappist the monk (talk) 21:54, 29 November 2014 (UTC)
 * —Trappist the monk (talk) 21:54, 29 November 2014 (UTC)


 * I was going to suggest using publisher instead of website so "Sutton Coldfield Natural History Society" is not italicized...
 * ...but I don't think that the title should be italicized either. GoingBatty (talk) 22:07, 29 November 2014 (UTC)
 * ...but I don't think that the title should be italicized either. GoingBatty (talk) 22:07, 29 November 2014 (UTC)
 * ...but I don't think that the title should be italicized either. GoingBatty (talk) 22:07, 29 November 2014 (UTC)


 * "Sutton Coldfield Natural History Society" isn't the name of the website; it isn't the title of any published work. So it shouldn't be italicized. As GoingBatty notes, it's the publisher. The individual web page title shouldn't be italicized as it's only a component of the un-titled website. The best that can be done with is probably something like:
 * In reality if I had to use a citation like this in an article using CS2 I would now have to use :
 * But this is not acceptable in my view; some way must be provided for users of to produce correct citations in cases like these. Peter coxhead (talk) 09:56, 30 November 2014 (UTC)
 * But this is not acceptable in my view; some way must be provided for users of to produce correct citations in cases like these. Peter coxhead (talk) 09:56, 30 November 2014 (UTC)
 * But this is not acceptable in my view; some way must be provided for users of to produce correct citations in cases like these. Peter coxhead (talk) 09:56, 30 November 2014 (UTC)


 * Ok, one might do this:
 * The banner image at the top of the left column identifies the target page as part of the Sutton Park web site. This I think is preferable to using sp.scnhs.org.uk which requires that readers decode it to understand it.
 * The banner image at the top of the left column identifies the target page as part of the Sutton Park web site. This I think is preferable to using sp.scnhs.org.uk which requires that readers decode it to understand it.
 * The banner image at the top of the left column identifies the target page as part of the Sutton Park web site. This I think is preferable to using sp.scnhs.org.uk which requires that readers decode it to understand it.


 * —Trappist the monk (talk) 12:08, 30 November 2014 (UTC)
 * But now you've just invented a website name – a plausible one certainly, but still your invention. The fundamental problem remains: I don't need to invent a website name if I use ; I do now if I use . I'm not persuaded that this is right. Peter coxhead (talk) 21:34, 30 November 2014 (UTC)


 * In some cases it will be more seriously wrong, with an error message. See Category:Pages with citations having wikilinks embedded in URL titles for 1200+ pages with problems, many of which (all the mathematics ones I checked) are caused by this change: they had a wikilinked title, and an external-linked (with url) contribution or chapter, but this change caused both links to be placed on the same text, causing multiple problems: (1) the wikilink coding is shown visibly rather than creating a wikilink, (2) the external link is in the wrong place, and (3) we get a red error message. My suspicion is that with so many of these errors, there are bound to be many many more harder-to-find errors where this change caused the link to be placed on the wrong text but there wasn't something else there causing a visible error. —David Eppstein (talk) 22:12, 2 December 2014 (UTC)


 * I moved your post out of Editor Peter Coxhead's post to here. Except for its position, your post is unmolested.


 * I acknowledge that there will be / are new errors that arise from this change. For the long term though, editors will benefit from the more rational way that CS1 and CS2 deal with the chapter (and alias) parameter set: chapter, chapter-url, trans-chapter, chapter-format all work together as a suite.  This in comparison to the promiscuous url of old which would abandon title and trans-title if chapter were present and not escorted by chapter-url.  No more.  Now, url, in keeping with its documented function, is bound to the title parameter set: title, script-title, and trans-title; it must compete with title-link.


 * —Trappist the monk (talk) 23:35, 2 December 2014 (UTC)
 * You've broken a lot of existing citations, broken the software I've been using to format citations, and (by using longer and harder-to-remember parameter names) made the templates more complicated and more difficult for human users to format. That doesn't seem like progress to me. Where is the discussion of and consensus for this big change? —David Eppstein (talk) 01:10, 3 December 2014 (UTC)


 * I am never silent about what I do. The discussion about chapter was at Help talk:Citation Style 1/Archive 6.


 * —Trappist the monk (talk) 01:21, 3 December 2014 (UTC)
 * So in other words it was never discussed at a talk page relevant to this template. —David Eppstein (talk) 16:56, 3 December 2014 (UTC)


 * you are quite right; the discussion has been overwhelmingly directed at CS1, and problems that the changes have caused to CS2 don't seem to matter. As I noted above, I cannot now achieve the same citation style with that I can with, yet I see no consensus for this change. Peter coxhead (talk) 22:26, 3 December 2014 (UTC)

Why is contribution level parameter now being ignored
I couldn't follow the discussion above, so this may be a duplicate, but the recent change to this template has been detrimental to some existing citations. Based on the documentation for this template, and former functionality, we could specify a contribution level parameter (ie chapter, contribution, entry, article, section) in addition to title & work, but that is no longer working, such as in this example:

It works slightly different if I just use the url parameter; previously that would default to the lowest level (ie contribution if defined; if contribution not defined then title), but the template is still not displaying the contribution level:

I assume this is a bug, but I'm not completely sure; can this be fixed if it is, and if this is somehow a new "feature", could someone explain the reasoning for this? — Asterisk *  Splat → 19:37, 2 December 2014 (UTC)
 * One possible workaround is to use series instead of work:
 * I note that website gives the same problems as work. Possibly this is the same issue as the one being discussed for cite conference at Help talk:Citation Style 1? —David Eppstein (talk) 20:08, 2 December 2014 (UTC)
 * I note that website gives the same problems as work. Possibly this is the same issue as the one being discussed for cite conference at Help talk:Citation Style 1? —David Eppstein (talk) 20:08, 2 December 2014 (UTC)
 * I note that website gives the same problems as work. Possibly this is the same issue as the one being discussed for cite conference at Help talk:Citation Style 1? —David Eppstein (talk) 20:08, 2 December 2014 (UTC)


 * needs to be able to figure out how you want the rendered citation to appear. It makes this determination based on the content of various parameters, most notably work (or any of its aliases: journal, newspaper, magazine, website).  When  templates use work or any of its aliases, the chapter parameters are ignored because periodicals and websites don't, as a general rule, have chapters.  When this occurs, title contains the title of the article or web page, work is the journal or newspaper or magazine or website name.


 * You could do this:
 * which gives:
 * which gives:


 * Editor David Eppstein's suggestion to drop work and use series 'works' because work when empty or omitted, tells that it is to render the citation as a book.  This looks ok but produces less that ideal metadata because the cited website is not a book nor is the not-book part of a series.


 * —Trappist the monk (talk) 20:52, 2 December 2014 (UTC)
 * Where did this dogmatic view of what periodicals can't possibly ever do come from, and who endorsed putting it into the template code? Periodicals absolutely do have two-level structures to some of their publications: individual pieces within a regular column, papers within a journal special issue, individual headlines within a news roundup, individual problems within a mathematics journal's problem column, etc. We need some way to refer to these things. —David Eppstein (talk) 21:07, 2 December 2014 (UTC)


 * Is Gospel Topics a regular column? If so, use department. --  Gadget850talk 21:38, 2 December 2014 (UTC)


 * Gospel Topics a reference work that is similar to, but not the same as an encyclopedia. The individual entry in the example I gave is "Baptisms for the Dead", the collection of entries is called "Gospel Topics" which is located on "LDS.org" and is published by the "LDS Church", and without all four of these elements the citation is incomplete. For many years we've been able to accommodate this structure using the contribution, title, work, and publisher parameters, and I still don't understand why suddenly this has to break &/or is no longer supported. — Asterisk *  Splat → 20:23, 3 December 2014 (UTC)


 * @Editor David Eppstein: Do you have real life examples? Editors can also use at.


 * —Trappist the monk (talk) 22:18, 2 December 2014 (UTC)
 * All of the examples I listed are real-life. The proceedings of SIGGRAPH are published as journal issues: how are we supposed to format the journal, the fact that it's the proceedings of a conference, and the title of an individual paper in a single citation without abusing series? We cite many problems from the problems column of the American Mathematical Monthly; the first one I found in a quick search was in Erdős–Mordell inequality, which just lists the problem number and the journal, but this will probably eventually cause problems for software like User:Citation bot because the actual title of the publication as listed by JStor is "Problems for Solution: 3739-3743" and we want to pick out one component from within that title. For many citations to subheadings within a roundup of news articles published within a journal that were all properly formatted a month ago when I added them and broken by this change, see ICTP Ramanujan Prize. Note also that in the last example, ignoring contribution means that the authorship is attributed incorrectly: each subheading of the "Mathematics people" article is individually signed (or not), so when these instances of citation have authors the author should be indicated as having written the contribution, but instead they are incorrectly indicated as having written the entire "Mathematics people" article. This incorrect attribution cannot be fixed by your suggested abuse of the at parameter. You might as well suggest we write the whole citation in the id parameter and not structure our data properly. —David Eppstein (talk) 01:17, 3 December 2014 (UTC)

department: Regular department within the periodical. Displays after work and is in plain text.

--  Gadget850talk 01:38, 3 December 2014 (UTC)
 * This doesn't address the fact that in the example I gave, I want to cite "Problem 3740" and the title of the article it appears in is something else, "Problems for Solution: 3739-3743". The department that you list is yet another level of subdivision intermediate between the article title and the journal. —David Eppstein (talk) 02:02, 3 December 2014 (UTC)
 * Sorry. No more ideas. Out. --  Gadget850talk 02:25, 3 December 2014 (UTC)
 * I do want to thank you for the department idea. It seems to be working well as a workaround on the Ramanujan prize examples. At least until a bot comes along and changes the title back to what the journal says the title is. Or until Trappist the monk gets it into his head that of course scientific journals can't have departments, disables that parameter for citations that have journal, and forces us to use newspaper instead as another workaround. (And if you think that example is ludicrous, see over on CS1 where suddenly cite conference is deprecated, conference proceedings citations don't exist any more, and they're all being made to be formatted as if they were encyclopedia articles.) —David Eppstein (talk) 17:46, 3 December 2014 (UTC)
 * Re SIGGRAPH: Perhaps follow the example I found here where there is a bibcode structure that looks like this:
 * @article{fakscm_metarep_sig14,
 * AUTHOR = "Noa Fish* and Melinos Averkiou* and Oliver van Kaick and Olga Sorkine-Hornung and Daniel Cohen-Or and Niloy J. Mitra",
 * TITLE = "Meta-representation of Shape Families",
 * JOURNAL = "Transactions on Graphics (Special issue of SIGGRAPH 2014)",
 * YEAR = "2014",
 * numpages = {11},
 * note = {* joint first authors}
 * }
 * Except for  which we don't support, that translates to this to which I've added url and format:
 * and that renders into this:
 * Alternately, using SIGGRAPH 2014:
 * Alternately, using SIGGRAPH 2014:
 * Alternately, using SIGGRAPH 2014:


 * From JSTOR:
 * 3740
 * Paul Erdös, L. J. Mordell and David F. Barrow
 * The American Mathematical Monthly
 * Vol. 44, No. 4 (Apr., 1937), pp. 252-254
 * Published by: Mathematical Association of America
 * Article Stable URL: http://www.jstor.org/stable/2300713
 * Using the above with 3740 p.&amp;nbsp;252 we get:
 * Taking away at and using Problems and Solutions and 3740 we get this:
 * Taking away at and using Problems and Solutions and 3740 we get this:


 * Similarly, department is an appropriate parameter for the ICTP Ramanujan Prize:


 * I notice that the department documentation says: Displays after work and is in plain text. This is incorrect.  It should read: Displays before work and is in plain text.  I'll fix that.


 * —Trappist the monk (talk) 12:49, 3 December 2014 (UTC)


 * A subheading of an article or department could be specified with at, since it is location information. However, the page number, if needed, would have to be put into at since at and page are mutually exclusive. Example
 * But this example illustrates that giving a department name without giving a title for that department doesn't work right. Jc3s5h (talk) 13:52, 3 December 2014 (UTC)
 * But this example illustrates that giving a department name without giving a title for that department doesn't work right. Jc3s5h (talk) 13:52, 3 December 2014 (UTC)
 * But this example illustrates that giving a department name without giving a title for that department doesn't work right. Jc3s5h (talk) 13:52, 3 December 2014 (UTC)


 * Why would you do that? The article title is "A better test than Turing", why put it in at? Suppose you wanted to cite the box at the end of the article, you might do this:
 * —Trappist the monk (talk) 15:11, 3 December 2014 (UTC)
 * —Trappist the monk (talk) 15:11, 3 December 2014 (UTC)


 * Magazines and journals don't have metadata that says "This is an article, it begins at this spot and ends at this other spot." They just arrange text and leave it to the reader to decide what is an article and what isn't. I decided "A better test than Turing" is a section within the News department. Chicago (14th ed. § 14.202 says "Titles of regular departments in a magazine are capitalized headline-style but not put in quotation marks." An example is given that has both a department title, Talk of the Town (Ink) and a title, "Isn't it Romantic?" But looking at the New Yorker website they seem to have one topic per issue in the Talk of the Town (Ink) department, while Spectrum has several topics per issue in their News department. So I don't think there is any clear-cut way to decide if one heading among several in a magazine department counts as an article. I'm sure it won't be obvious to editors using the Citation template.


 * More generally, it may be convenient to name a heading with at that is unambiguously a section heading, and certainly not an article; this could come up with any kind of article, not just departments. Jc3s5h (talk) 15:45, 3 December 2014 (UTC)


 * At the New Yorker TOC for 1 December 2013, scroll down to The Talk of the Town. There, it would seem, are multiple 'things' with titles so:
 * Alas, no heading in that article. The 'Ink' tag appears to be a semi-regular 'category'; the tag 'Comment' is quite regular, but then they use tags like 'Bob's Boys' from the 8 December 2014 issue.  What are we to make of that?  Bob's Boys: All natural?
 * Alas, no heading in that article. The 'Ink' tag appears to be a semi-regular 'category'; the tag 'Comment' is quite regular, but then they use tags like 'Bob's Boys' from the 8 December 2014 issue.  What are we to make of that?  Bob's Boys: All natural?


 * —Trappist the monk (talk) 16:44, 3 December 2014 (UTC)
 * For me, which heading to use the title of a journal article is usually unambiguous: it's what the journal's web page for the article lists as a title, and what appears in the table of contents. The supposed formatting for the solution to problem 3740 above is ok, for instance, because that's what it lists, but it again misses the point, because the reference that I was having trouble formatting is not that one (where the solution was published) but the earlier one where the problem was posed. And at is usable as a catch-all in case the existing parameters fail to work, but doesn't seem to have any way of having a url attached to it, and the inability to use both at and pages is again problematic. —David Eppstein (talk) 16:49, 3 December 2014 (UTC)


 * Yeah, I've occasionally wondered if there isn't some way to allow either page or pages and at. I've never reached a determination.  at can be linked.  This is the previous example with 3740 p.&amp;nbsp;252
 * I guess I'm not understanding what it is you're looking for. How would citing problem 3740 be different from citing solution 3740?
 * —Trappist the monk (talk) 18:00, 3 December 2014 (UTC)
 * They are two different publications in two different issues of the journal. Problem 3740 (the one in which the problem was posed) is part of a larger set of problems; the set of problems has a jstor id (or doi) and a title referring to the whole set, but I want to actually refer only to one problem within that set. (In particular, I want to give authorship credit to the person who posed that problem, and not to the other people listed as aggregate authors of the set of problems). The solution to problem 3740 is a separate citation for a later issue of the journal where a solution was published; it has (or could have) different authors than the original problem statement, it certainly has different volume/issue/page data, and in this case is less problematic because it happens to be listed by jstor as its own separate publication. —David Eppstein (talk) 21:08, 3 December 2014 (UTC)
 * They are two different publications in two different issues of the journal. Problem 3740 (the one in which the problem was posed) is part of a larger set of problems; the set of problems has a jstor id (or doi) and a title referring to the whole set, but I want to actually refer only to one problem within that set. (In particular, I want to give authorship credit to the person who posed that problem, and not to the other people listed as aggregate authors of the set of problems). The solution to problem 3740 is a separate citation for a later issue of the journal where a solution was published; it has (or could have) different authors than the original problem statement, it certainly has different volume/issue/page data, and in this case is less problematic because it happens to be listed by jstor as its own separate publication. —David Eppstein (talk) 21:08, 3 December 2014 (UTC)


 * Comment I have to say it is ludicrous that we can no longer specify something as basic and as essential as a chapter in a long document. Take Motion picture rating system for example, which specifies a chapter in a 30-page document (which is divided by chapters). I am sure there are hacks to get around this, but i) that's a lot of unnecessary fixing and ii) you should be able to specify the chapter in the chapter parameter. Betty Logan (talk) 17:04, 3 December 2014 (UTC)


 * You can, but that citation isn't clearly citing the document. The chapter and title belong to the document in the window and the url points to the web page that has the document and a video and a twitter feed and some other stuff.


 * Consider these, first as and then as  (a redirect to ):
 * or your example citation, unchanged except that instead of it is :
 * —Trappist the monk (talk) 18:00, 3 December 2014 (UTC)
 * or your example citation, unchanged except that instead of it is :
 * —Trappist the monk (talk) 18:00, 3 December 2014 (UTC)
 * —Trappist the monk (talk) 18:00, 3 December 2014 (UTC)

arxiv parameter
Starting in 2015 (ie, after 1412.xxxx), serial numbers are 5 digits long (e.g., 1501.01586). Check code complains about 5 digit numbers. Glrx (talk) 04:37, 11 January 2015 (UTC)
 * (Right)
 * (Wrong)


 * Fixed in sandbox; see Help talk:Citation Style 1/Archive 7 --  Gadget850talk 08:24, 11 January 2015 (UTC)

Date range error for journal issues
The cite journal template reports an error if I enter the date for a journal issue covering a period of time. For example April-June 2014. Is there away around this without sticking the period in the "issue" field?

For example:



Produces the following error text:
 * "Check date values in: |date= (help)"

I don't particularly want to put "issue=50 (July-August)".

Thanks, Simon Burchell (talk) 12:46, 12 January 2015 (UTC)


 * As noted on the help page, you are using a hyphen instead of the en dash.




 * --  Gadget850talk 13:11, 12 January 2015 (UTC)
 * Thanks - I had tried playing with en-dashes a while back until I gave up in exasperation. Ideally, these things should be easy to just type in. I am more interested in writing articles than in exploring obscure items of syntax - the example is helpful. Many thanks, Simon Burchell (talk) 13:22, 12 January 2015 (UTC)


 * You can use the charinsert tool below the edit window. When 'insert' is selected, the first character is an en dash. --  Gadget850talk 16:48, 12 January 2015 (UTC)
 * So it is... thanks again! Best regards, Simon Burchell (talk) 16:55, 12 January 2015 (UTC)

origyear seems to suppress year
As seen here. EEng (talk) 12:11, 18 April 2015 (UTC)


 * There are two versions of that Three Centuries of Harvard citation. One uses origyear, the other does not.  This one does:


 * —Trappist the monk (talk) 12:24, 18 April 2015 (UTC)
 * Oops. EEng (talk) 12:43, 18 April 2015 (UTC)

RFC link
I would expect including |rfc=792 to produce a internet link to the RFC in the footnote. It does not. Including RFC 792 in body text does magically create a link. ~KvnG 16:00, 17 January 2015 (UTC)
 * Nevermind. I see that the link is on the title, not the RFC number. ~KvnG 16:03, 17 January 2015 (UTC)

Citation/patent PS parameter
Citation/Patent is called with this default expression to the PS parameter: |PS = with 4 parameters to the #if, one too many, apparently. Maybe it should be an #ifeq. Attardi (talk) 04:48, 22 January 2015 (UTC)

Anchors
I've recently been working on Tir national and am trying to tie up the citations. One problem that has surfaced is the lack of authors for web sites, particularly those in a foreign (to me) language. In order to get an anchor for short form footnotes, I have simply forced the ref to be the title: ref = CITEREFTir_National for example. In the absence of an author the title displays first and so both the electronic linkage and the human's view agree: ^Tir National looks right.

The problem comes with long names. A title such as: "Ancienne caserne Prince Baudouin, dite également caserne Dailly" can hardly be called SHORT form! Basing the anchor on a subset works electronically but is confusing to the reader. What I'm trying is putting (in this case) "History:" before the citation, then using ref = CITEREFHistory and. I did look at the display options, specifically author-mask, but that only works when there is a real author. Advice/comments please! Martin of Sheffield (talk) 12:08, 22 January 2015 (UTC)
 * The advantage of is the ability to easily reuse a source with different page numbers. With only two references in this article and one to a website I would recommend basic . --   Gadget850talk 12:25, 22 January 2015 (UTC)
 * I understand your point but I'm afraid that personally I disagree. Starting an article with one form and changing it later if it expands can be a right pain, it is far better to do the job properly from the start.  I've been expanding this article recently and there is a little more to come.  Secondly, I find that lengthy footnotes buried in the middle of the text make reading it harder which enourages mistakes (such as getting the name= ever-so-slightly wrong) and discourages other editors.  Thirdly, it keeps a common interface with sfn and efn.  All my personal opinion of course; I'm only a lowly Yeoman not one of the multikiloedit superstars!  Leaving preferences apart though, is there a neater way to anchor a page with a long title and no obvious author?  I may post the original question on sfn and see if people there come up wih a solution.  Thanks, Martin of Sheffield (talk) 14:23, 22 January 2015 (UTC)
 * On the few occasions that I have neither author nor editor, I've put together something from either the title or the publisher. You can see both being done at NBR 224 and 420 Classes - the nine refs to SLS 1970 and the single ref to Gradient Profiles 2003. In a worst-case situation, I have used Anon in the so that  works.  -- Red rose64 (talk) 14:49, 22 January 2015 (UTC)
 * Thanks. That's the sort of thing I did at first, I'm probably being over picky trying to improve though.  Unless you actually click on the citations there is no obvious link from SLS to "Locomotives of the North British Railway", "Gradient Profiles" is a bit more obvious.  Compare to "Highet 1970 p. 88" which rather obviously is "Highet, Campbell (1970) ...".  Nice article BTW, though not I fear one of the more elegant locos to look at. Martin of Sheffield (talk) 15:04, 22 January 2015 (UTC)
 * The infobox image is a scan of a photo that I found in a 1987 facsimile reprint of a book published in 1927, itself compiled from articles in The Engineer which appeared between 1925 and 1927. I suspect that the 1987 edition didn't use the original negative, although the 1927 one probably did. So it's at least a third-generation copy, and not of the best contrast or quality. -- Red rose64 (talk) 16:13, 22 January 2015 (UTC)

Cite conference chapter bug
About the Help:CS1_errors bug. I understand it will be solved in the next update. That's fine. For until then, I propose to create a temporal module that serves (6k transc's) to prevent this bug appearing. Setup:
 * Create module:citation/CS1/conferenceTemporal.
 * Has the current code from module:citation/CS1/sandbox (ie latest stable /sandbox code, bug-free).
 * uses.

Background: this is why a bug is called a bug. At the moment people are bringing Astatine to FA level, and explaining this is taking too much editing energy from all of us. -DePiep (talk) 11:54, 5 February 2015 (UTC)


 * Technical issues should never affect FA reviews, but the article is not at FAC yet, so this is not an immediate problem. The bug has been resolved in the sandbox and will be deployed when it all other changes have been tested. And I don't see an error on the article page (other than obsolete HTML in the infobox template). --   Gadget850talk 12:08, 5 February 2015 (UTC)
 * I did not say it prohibits FA, I said it takes away editor's energy. To that you have not replied. You saying "it will be deployed" is what I already write + my opinion on that. (And are you saying 'I did not see it so there is no bug'?). -DePiep (talk) 12:23, 5 February 2015 (UTC)


 * The comment on the talk page "Okay, I see. So I think I should leave it as is and wait for the bug to be fixed." leads me to infer that this is not a major problem. Astatine currently shows no CS1 errors. --  Gadget850talk 12:28, 5 February 2015 (UTC)
 * 1. That "Okay" was by one editor. Then another one came along - starting it all over. q.e.d. (Did you really spend time on this search? And then you did not see the error being explicitly pointed to in that same page?). 2. No one said "major". Another needless deviation. 3. My example stands for an unknown number of cases. A bit too simple to "kill the example, and everything is gone". 4. Astatine does have this error. And when it is removed by editing into a different solution, that too supports my point (plus that it could mean a lesser solution was chosen). 5. I get the impression that you don't like my question. To spend so much time on denying the issue, I don't see why. You leave it to me to read between your lines "I don't like it". -DePiep (talk) 13:00, 5 February 2015 (UTC)


 * I just observed somebody else changing  to   on a  template (  is transformed to   before   is ignored…).  It would seem better to fix this in the code, than to have all these people running around kludging parameters to avoid showing apparent citation errors.  If it's a warning (rather than a fundamental error) it would probably be preferable to silently ignore it until fixed rather than generating big red errors.  —Sladen (talk) 12:38, 5 February 2015 (UTC)


 * Revert and note that the issue is being fixed per the linked help page. Do we want to rush the deployment? --  Gadget850talk 13:27, 5 February 2015 (UTC)
 * Looking at the change it seems proper. Journals don't have chapters but the do have departments. The issue in question here is with cite conference. --  Gadget850talk 13:36, 5 February 2015 (UTC)


 * Ideally a combination: (1) suppress the display of "errors" for unrecognised non-essential parameters, then at a later date (2) ship the additional support for  when it's ready and tested.  This would, I hope, solve both the immediate temptation for editors to 'fix' unnecessary errors-that-aren't-errors, and also allow deployment of the expanded codebase without rushing.  Would that work?  —Sladen (talk) 13:38, 5 February 2015 (UTC)  Yes, but magazines have sections…


 * So don't show the error but silently don't show chapter? That is exactly how we got into this situation. cite conference never supported 'chapter': see the template history. Apparently editors used 'chapter' but never noticed that it did not work. This issue arose only when we started adding the error checking and the errors showed up. And 'department' works nicely for sections in journals, but again we aren't disscussion that template. --  Gadget850talk 13:49, 5 February 2015 (UTC)
 * I think you are wrong to call citations formatted this way "errors" — they are only marked as errors because the template declares them to be so, not for some intrinsic reason — and your description of how these came to occur does not match my experience. I think most of the errors with cite conference were caused by using Title of a paper in a conference proceedings and Title of the proceedings which used to work but at some point became broken by treating these two parameters as if they were chapter/title. The workaround is to use conference in place of booktitle. And department works for some forms of hierarchy within journals, but not others; e.g. a journal I subscribe to (Bull. AMS) regularly has special articles that have their own editor (who is not the journal editor) within which there are sub-articles that have their own separate authors. It used to work in citation (the template we should actually be discussing here) to cite these things using title for the umbrella article and contribution (or chapter if you prefer) for the sub-articles, but now it does not. Another example, bringing it back to cite conference: some conference proceedings are published as issues of journals. It used to work to cite papers in such proceedings using contribution/title/journal (matching the normal parameters one would use for a paper in a proceedings and a title within a journal) but now it does not; the workaround, using series in place of journal, does not format the journal title in the same way.

This "conference-chapter" association is ambiguous, and needs a closer look. Reports presented at a conference are often published as proceedings, and I believe I have seen one such work that had chapters in the same manner as a book or such. However, the conferences themselves do not have chapters except as a subgroup of a larger organization, often the sponser of the conference. I don't know that there has been any such usage, but it seems like a strong possibility. ~ J. Johnson (JJ) (talk) 21:47, 5 February 2015 (UTC)
 * Why do yo all reopen the discussion? I was just proposing to remove a bug that was to be removed anyway. The curious thing is that now my OP was not picked up seriously, and an archive page is retyped. -DePiep (talk) 22:15, 5 February 2015 (UTC)

The request above will be addressed in a bug fix that is scheduled to go live by February 15. See Help talk:Citation Style 1 for further details. – Jonesey95 (talk) 22:46, 7 February 2015 (UTC)

Solicited comments to an article
Sometimes a journal editor will ask authorities to comment on an article. For example, Marian Rejewski wrote an article, and it was published with afterwords by Good and Deavours. The DOI for the article includes both the article and afterwords. Is there a good way to handle these afterwords?


 * I. J. Good and Cipher A. Deavours, afterwords to: Marian Rejewski, "How Polish Mathematicians Deciphered the Enigma", Annals of the History of Computing, 3 (3), July 1981. http://chc60.fgcu.edu/images/articles/rejewski.pdf

Glrx (talk) 21:47, 14 February 2015 (UTC)

coauthors
Where has it been agreed that coauthors ought to be deprecated? -- PBS (talk) 21:08, 13 February 2015 (UTC)
 * Why do you keep asking this question? -- Red rose64 (talk) 14:24, 14 February 2015 (UTC)
 * Perhaps because it hasn't been satisfactorily answered? Peter coxhead (talk) 17:51, 14 February 2015 (UTC)
 * Just found an archived discussion; maybe it could shed some light? &rarr; Template talk:Citation/Archive 6. --User:Ceyockey ( talk to me ) 18:18, 14 February 2015 (UTC)
 * Let me be clear that I think it's better not to use coauthors, so I would support it being deprecated, but I'd like to see where the consensus was reached to do this. The discussion linked above assumes it is deprecated, so doesn't answer PBS's question. Peter coxhead (talk) 20:52, 14 February 2015 (UTC)

My concerns on this are two fold. The first is that the programming tail seems to be wagging the dog. For example where is the process that decided this is the correct course of action? The second is practical coauthors is an extremely useful for separating out those authors one wishes to include in a short citation and those one does not, so there is a practical use for coauthors so why ought it to be deprecated as those who do not want to use it do not have to. Now it may be that after a formal process of deprecation it is decided to ditch coauthors. But deprecation and the introduction of new parameters ought to be done through a process which is easily available to editors who are not familiar with the underlying programming languages/scripts. -- PBS (talk) 21:08, 14 February 2015 (UTC)
 * , you might be able to find a discussion on an archive page within one or two months prior to this edit, which added coauthors to the deprecated parameters category page. – Jonesey95 (talk) 00:38, 15 February 2015 (UTC)
 * Can't find anything that looks like a consensus. -- PBS (talk) 14:58, 16 February 2015 (UTC)

authorformat smallcaps error
The authorformat=scap, producing smallcaps, is now producing an " Invalid |name-list-format=scap" error across a large number of articles. Anyone have any idea why? Simon Burchell (talk) 15:04, 15 February 2015 (UTC)


 * author-format and editor-format have been obsoleted in favor of name-list-format. The small caps formatting has been removed in compliance with Manual_of_Style/Capital_letters.


 * —Trappist the monk (talk) 15:30, 15 February 2015 (UTC)
 * Was this actually discussed anywhere, or is it just another mass change being forced across all articles? It seems drastic. Simon Burchell (talk) 15:33, 15 February 2015 (UTC)
 * This has been in the MOS for years; we're just catching up here. I think the MOS talk page would be the proper place to debate making an exception for author name formatting in references, but that exception is not there now. —David Eppstein (talk) 18:05, 15 February 2015 (UTC)
 * MOS does not apply to citations so it is invalid to use a passage in the MOS as justification for a citation style change. Jc3s5h (talk) 15:36, 16 February 2015 (UTC)
 * what's the basis for this assertion? It's an important question because seveal changes to the citation templates, including forbidding open-ended date ranges, have relied on guidance in the MOS. Peter coxhead (talk) 16:55, 16 February 2015 (UTC)
 * A discussion started at Module talk:Citation/CS1/Archive 8 in which it was decided that CS1 would adopt the date formats from WP:MOSNUM. I should acknowledge that there is considerable disagreement about whether MOS can control citations; the compromise so far has been to word WP:MOS, WP:MOSNUM, and WP:CITE so they don't contradict each other. Jc3s5h (talk) 17:25, 16 February 2015 (UTC)
 * Your second sentence suggests that your comment "MOS does not apply to citations" isn't really correct. Peter coxhead (talk) 17:43, 16 February 2015 (UTC)
 * I just find the constantly-changing parameters tiresome; the citation template isn't at all stable. I am far more interested in writing articles, than constantly running to catch up on parameter usage. I think I'll just stop using the citation templates in future. Thanks for replying, Simon Burchell (talk) 19:05, 15 February 2015 (UTC)

, I think you have a script that removes authorformat=scap (and its variations, if there are any) from articles. Maybe it would be worth running it through the deprecated parameter category periodically for the next few weeks until the job queue catches up. That would reduce the amount of easy-to-fix red ink seen by some of WP's most prolific contributors. – Jonesey95 (talk) 06:52, 16 February 2015 (UTC)


 * I do and have begun to do that.


 * —Trappist the monk (talk) 12:26, 16 February 2015 (UTC)


 * @David Eppstein The MOS is a guideline and as such changes that affect a lot of pages needs to be discussed to see if there is a consensus for a change (guidelines are far from mandatory). @Trappist the monk The templates are not at all popular with many editors and the last thing that those who like to use them need is for the interfaces to be change on a regular basis. I think you need to take Simon Burchell last comment to heart. -- PBS (talk) 15:04, 16 February 2015 (UTC)


 * How long will it take to remove these by script? I spent half an hour trying to figure out why an article is red all over. I don't see the edit or understand it, and until I came to the talk page, I could find no explanation, even though the template mentions another parameter no longer in use. Are there other planned but unexplained and invisible-to-users changes that will be happening? How do I figure out the changes, just see red and assume something changed and delete it?(I know I can just text edit it, but better to know in advance not to use a template that will change a lot.) MicroPaLeo (talk) 05:32, 26 February 2015 (UTC)


 * By script, not very long. But, see the discussion at Wikipedia talk:Manual of Style/Capital letters


 * —Trappist the monk (talk) 12:26, 26 February 2015 (UTC)

Chapter
Some change has been made so that chapter field is being ignored, and producing an error. eg



I can't see any good justification for this, it has made many good references now produce errors, and even this page's own documentation shows errors. Was there a reason for this change, or unforced error? Please change this back/explain. Thank you.83.100.174.82 (talk) 11:21, 13 January 2015 (UTC)


 * I don't read French, so it is difficult to pick out this example. What is "La traction électrique férroviaire" the title of?


 * As the help page notes, if you use journal or an alias, then chapter is ignored. Journals have articles, not chapters.


 * And the example on the template page is not well formed:


 * It looks like this is trying to cite a section within a news article, specifically an affidavit, which forms the bulk of the article. --  Gadget850talk 12:19, 13 January 2015 (UTC)


 * I think that the original example really refers to some kind of report rather than a journal article. As I understand it, La revue 3E.I is a series of reports, of which the one numbered 19 is titled La électrique férroviaire, then pp. 23–34 within it are the chapter/section "La BB 36000, une locomotive 'passe frontières'". If this is correct, it should be set it up like this:
 * Peter coxhead (talk) 12:39, 13 January 2015 (UTC)
 * Peter coxhead (talk) 12:39, 13 January 2015 (UTC)


 * The issn suggests that La Revue 3 E I : enseigner l'électrotechnique et l'électronique industrielle [Review 3 EI: teaching electrical engineering and industrial electronics] (google translate) is a periodical. The linked web page, apparently not associated with La Revue 3E.I lists the title as "La BB36000 la locomotive 'Passe - frontières'" with a date of December 1999, issue 16. That same web page lists 'Source' as "La traction électrique férroviaire", December 1999 but issue 19.  Which issue number is correct?


 * Is it possible that 'La traction électrique férroviaire' [Railway electric traction] is a department in the journal? If so, then the citation can be rewritten:
 * If the source being cited is the website thierry-lequeu.fr, then the citation reduces to this:
 * If the source being cited is the website thierry-lequeu.fr, then the citation reduces to this:


 * —Trappist the monk (talk) 13:18, 13 January 2015 (UTC)

Is is possible that this undiscussed change be reverted. It's not a single problem - take a look at the references in the high importance page USB - I count 4 examples of this error. I am seeing this in many places.83.100.174.82 (talk) 18:18, 25 January 2015 (UTC)
 * I have fixed those poorly-formatted citations, along with other citation errors in the USB article. Thanks for bringing it to our attention. – Jonesey95 (talk) 18:48, 25 January 2015 (UTC)
 * It's not just poorly-formatted citations. I continue to be incredibly frustrated by this damage to the citation templates by User:Trappist the monk and his refusal to undo it even after having its bad effects pointed out. I just had to hack around this problem for several previously-working citations on Andrew M. Gleason, in which an obituary article published in a journal had different subsections with different authors (writing about different aspects of the subject's life). Instead of formatting it in the obvious way with The title of the obituary and The subsection with different authors I apparently had to redo all of those to use title/department because Trappist somehow became convinced that it would never be a good idea to cite smaller units within a journal article. —David Eppstein (talk) 05:42, 26 January 2015 (UTC)
 * the problem is that the workaround results in an incorrect citation, since it uses e.g.  when clearly "Andrew M. Gleason 1921–2008" is not a department within a journal. Fortunately the department doesn't appear in the metadata generated for such a citation (at least at present). Let's hope no-one decides to error-check the values of department... I'm frustrated by this and some of the other changes made to the citation templates, and find myself converting some existing uses to plain text to avoid "error" messages for what I regard as perfectly correct citations. I wonder if converting to plain text may be better in these cases too. Peter coxhead (talk) 18:09, 26 January 2015 (UTC)


 * Is this going to be fixed or not? This is a problem literally everywhere - the same problem has been introduced into cite journal eg Help_talk:Citation_Style_1
 * eg




 * These uses are legitimate, and I know that it worked before for both CS1 and CS2. Try to get someone to understand that their good faith edits broke something and it would be really cool for them to undo those edits.83.100.174.82 (talk) 13:16, 2 March 2015 (UTC)


 * I would use cite book for any source with an ISBN.




 * I have been working my way through the "chapter= ignored" category, and I have found that the majority of the articles are easy to fix. There are some instances of that I haven't figured out how to fix, so I have left them for further investigation. – Jonesey95 (talk) 20:00, 2 March 2015 (UTC)
 * There is a small but nonzero number of instances that are impossible to "fix" because they are not broken. It is very frustrating that the maintainers of these templates have refused to even admit this. The only resort I have found to work in some cases is to move parts of the citation information out of the template and format it manually instead. —David Eppstein (talk) 21:42, 2 March 2015 (UTC)
 * I believe you. I think that I have come across templates that are attempting to cite a section of a long journal article. I can't locate one at the moment, but that seems like a legitimate use of the template that should be available to editors.


 * Based on my reading of the documentation and your example above, it looks like is also unable to cite a section of a long journal article, but that is not entirely clear to me. The contribution parameter and its function are not currently listed in the documentation for, except in an example (copied above) that currently generates an error message (and which presumably used to work fine).


 * It might be best to start a new discussion at Help talk:Citation Style 1 to ask how to create a citation with the title of a work, the title of a chapter or article within that work, and the title of a section within that chapter or article. A clear-cut and specific example from an actual article would be useful to show the need for such a citation. It seems to me that, , and should all support that sort of citation.


 * BTW, I have come across plenty of templates using chapter and journal, which are rightly flagged by the module as erroneous. I have been changing chapter to title to fix those. – Jonesey95 (talk) 22:32, 2 March 2015 (UTC)
 * If you're collecting examples of different types of journal article citation that need both a title and a lower-level subunit within the title, here are three:
 * This diff (a self-citation) is to a section within a published journal article. The author's online copy of the article has separate urls for each section, so the recent change to make url= go to the title and not the contribution also broke the placement of the link. I could find no good way to properly format this one using the now-broken citation template and resorted to partial hand formatting, which of course completely loses the metadata.
 * Many of the citations in Andrew M. Gleason are to articles published within a special section of the Notices of the American Mathematical Society; this section had its own editors (different from the journal editors). Prior to this change to the citation template, they were cited as contribution=individual article (with author referring to the author of each article within the section) and title=the title of the special section (with editor referring to the editors of the special section). That has been changed to use department= for the name of the special section, but that generates the wrong metadata because it is not really a department, and the wrong formatting because the editors are placed at the level of the individual articles rather than the name of the whole special section.
 * The proceedings of the Eurographics conference are regularly published as special issues of Computer Graphics Forum, a journal. So it used to be possible to use contribution=Name of paper in the proceedings, title=Eurographics, journal=Computer Graphics Forum. But now the best way to format this is to pretend that Computer Graphics Forum is a book series instead of a journal, but this generates the wrong formatting as well as the wrong metadata. See for example the (not templatized) first entry in the selected publications section of Bernhard Preim.
 * —David Eppstein (talk) 22:55, 2 March 2015 (UTC)


 * I am also seeing this issue on Explorer II. The revision is inconsistent with the documentation for this template, so one or the other needs to be addressed. Praemonitus (talk) 21:41, 18 March 2015 (UTC)


 * Write DeVorkin like this:
 * I changed the template to, changed work to conference, added cs2 to be stylistically consistent with the other citations in the article, and removed 3642297188 because that refers to Dictionary of Minor Planet Names (Special:BookSources/3642297188).
 * For Froes, et al.: remove Magnesium Technology: Metallurgy, Design Data, Automotive Applications which, if one is to believe the Google books cover image, is not the title and is superfluous to the citation because work, and its aliases, force to treat the citation as a journal citation not as a book citation:
 * —Trappist the monk (talk) 22:31, 18 March 2015 (UTC)
 * For Froes, et al.: remove Magnesium Technology: Metallurgy, Design Data, Automotive Applications which, if one is to believe the Google books cover image, is not the title and is superfluous to the citation because work, and its aliases, force to treat the citation as a journal citation not as a book citation:
 * —Trappist the monk (talk) 22:31, 18 March 2015 (UTC)
 * —Trappist the monk (talk) 22:31, 18 March 2015 (UTC)
 * —Trappist the monk (talk) 22:31, 18 March 2015 (UTC)
 * —Trappist the monk (talk) 22:31, 18 March 2015 (UTC)

Updates and improvements to Citation Style 1 module code
Some updates and improvements to the Citation Style 1 module code are scheduled for April 18–19. See this discussion for more details. – Jonesey95 (talk) 13:49, 12 April 2015 (UTC)

pinpoint parameter
What in the world is this? It's used e.g. in. EEng (talk) 02:35, 4 May 2015 (UTC)
 * You can see where it appeared in this diff. It does not create an error message because the named ref that is actually rendered is in in the reflist. The contents of the other instances of the named ref are ignored.


 * pinpoint is a supported parameter in Cite court. The citations used Cite court earlier in the article's development; it looks like the editor was trying to cite page numbers but did not succeed. If you want to change them to use Harvard-style refs with page number references, I'm sure it would look nice. Not that my friend EEng would know anything about Harvard.... – Jonesey95 (talk) 02:56, 4 May 2015 (UTC)
 * You're just jealous because you went to Yale. Thanks for the explanation -- I couldn't find pinpoint in the citation doc so I thought maybe there was some strange keyword magic going on. EEng (talk) 05:39, 4 May 2015 (UTC) BTW, perhaps you'd enjoy Jonesey.

Dead link notices in format parameter
I don't know if this is a citation bot error or a template error but when Citation bot used to add Template:dead link into the format parameter like here, it would insert the article into Category:ALL ARTICLES WITH DEAD EXTERNAL LINKS and other such articles. Moving it out (like here) moved it into the proper Category:All articles with dead external links. I can't figure out in which template is it going wrong. -- Ricky81682 (talk) 13:29, 18 May 2015 (UTC)
 * It was wrong, then and now, for Citation bot to add to format.  format is to be used to identify the electronic file format of the source pointed to by the content of url.  It is  that adds the category, not Citation bot.  The normal output of, from your example page, is:
 * The current version of Module:Citation/CS1 unconditionally capitalizes everything in format. Because  is processed before, Module:Citation/CS1 capitalizes all of the above which is where the redlinked category arises.  At the next update to Module:Citation/CS1, the content of format will no longer be capitalized.
 * The current version of Module:Citation/CS1 unconditionally capitalizes everything in format. Because  is processed before, Module:Citation/CS1 capitalizes all of the above which is where the redlinked category arises.  At the next update to Module:Citation/CS1, the content of format will no longer be capitalized.


 * 1) none of that stuff in format in your example citation should be there because it is in violation of the parameter's definition
 * 2) Citation bot was in the wrong to place it there
 * 3) Module:Citation/CS1 improperly capitalizes the content of format (this has been fixed in the sandbox version)
 * —Trappist the monk (talk) 13:56, 18 May 2015 (UTC)
 * Move the template outside the, but leave it inside the . Thus, it will be something like   or similar. -- Red rose64 (talk) 17:53, 18 May 2015 (UTC)
 * Ok, it didn't match with the documentation for the template at all so that confirms it. I'll make a bug report to the bot in case it does those citations again. -- Ricky81682 (talk) 20:29, 18 May 2015 (UTC)

Request to add new parameter.
Please add a new "reviewed" parameter to be used on all cite templates. When not set, the default will be yes, and the template behaves as normal. When the parameter is set to no, a text is added in an appropriate location, stating the citation needs to be reviewed. This feature will be used by Cyberbot in the foreseeable future. I would propose an X to Y change, but, I'm not very good with Lua.—cyberpower  Chat :Online 16:17, 6 June 2015 (UTC)

Oppose. The concept of reviewing templates has never even been discussed, never mind approved. User:C678 has not posted the location of any request for bot approval of this proposed bot. Concerned. Although discussed at the village pump (proposals), I don't recall seeing a pointer to the discussion in any of the citation-related talk pages, such as this one, Help talk:Citation Style 1, or Wikipedia Talk:Citing sources. Jc3s5h (talk) 16:44, 6 June 2015 (UTC), revised 17:35 UT.
 * Please see this discussion which discusses the support for the concept of such a bot and how modified citations should be marked for review.—cyberpower  Chat :Online 17:07, 6 June 2015 (UTC)
 * To respond to your revised statement, the proposed change makes the most sense for simplicity and ease of use. The current proposed change makes no change to the current output on already used templates, thereby no damage.—cyberpower  Chat :Online 17:39, 6 June 2015 (UTC)
 * I have removed the edit-protected template, since it is for uncontroversial changes only, and this is a feature request that will require some discussion, development, and testing.


 * The Citation template uses Module:Citation/CS1 to render citations. If it is your intent to have this reviewed parameter added to all of the CS1 templates (,, etc.), I recommend that you post this request at Help talk:Citation Style 1. – Jonesey95 (talk) 20:54, 6 June 2015 (UTC)
 * Thank you. I will hold off until the bot can get its approval before posting again at CS1, since you yourself said it's for uncontroversial changes, and my seemingly uncontroversial edit request seems to have raised some concern.—cyberpower  Chat :Online 08:15, 7 June 2015 (UTC)
 * I don't know how the bot can be tested successfully without this change being made to the module. Adding yes to an existing citation will cause a red error message to appear, since reviewed is an unsupported parameter. – Jonesey95 (talk) 14:20, 7 June 2015 (UTC)
 * Really? I thought it would it get ignored, like normal templates.—cyberpower  Chat :Online 15:00, 7 June 2015 (UTC)
 * Citation templates are checked for all sorts of errors. See Help:CS1 errors for a list of red error messages generated by citation templates. – Jonesey95 (talk) 15:22, 7 June 2015 (UTC)
 * I did not know that. I just learned something new.  If the bot gets approved for a trial, I'll submit the edit request in the appropriate location to have the parameter added.  Thanks for your help.—cyberpower  Chat :Online 16:22, 7 June 2015 (UTC)

I'm not quite sure what this would achieve exactly. It this not what verify source is for already? Headbomb {talk / contribs / physics / books} 19:44, 7 June 2015 (UTC)
 * Sort of. The idea is that the bot-updated archived page might have different information than the one used for sourcing and thus needs review. I doubt that throwing this special case (which would be pretty large given all the URLs which have archived copies) into the general template is the best solution, though. Jo-Jo Eumerus (talk) 21:37, 7 June 2015 (UTC)
 * I could also use the verify source template, but it's easier to parse a parameter within the template itself, than to look for one that applies to the cite template. Either way I will make it work.  There's still plenty of time to figure out I should go about, as the bot is still in early development.—cyberpower Chat :Limited Access 05:27, 8 June 2015 (UTC)


 * Support "silent" version. We've long needed a parameter like this, as well as one that indicates whether a source is primary/secondary/tertiary for the claim in question, but operating without any visible output, and just used by categories, bots, gnomish editors, etc.  — SMcCandlish ☺ ☏ ¢ ≽ʌⱷ҅ᴥⱷʌ≼  20:58, 30 July 2015 (UTC)

Default for 'postscript'
Wasn't the default changed to a period at some point? I've been using the citation template on this basis for a while now, only to find out it's been put back to a null character. Praemonitus (talk) 15:47, 31 July 2015 (UTC)
 * The documentation has said the following since 2012: "postscript: Controls the closing punctuation for a citation; defaults to none." That doesn't mean that the documentation has always been correct, but that's what it says, at least. – Jonesey95 (talk) 16:01, 31 July 2015 (UTC)
 * The cite journal template and other Citation style 1 templates end with a period; that's one of the differences between them and the Citation style 2 produced by this template. Maybe that's what you're thinking of. —David Eppstein (talk) 16:19, 31 July 2015 (UTC)
 * It sounds like I was mistaken then. That's pretty disappointing... but thanks for the response. Praemonitus (talk)

Relevant discussions about parameters at Help Talk:CS1
There are multiple discussions at Help Talk:CS1 that may affect the rendering of this template. It is recommended that people who watch this talk page also watch that one. – Jonesey95 (talk) 18:01, 8 August 2015 (UTC)

Style: et al.
When "et al." is inserted at the end of an author list (right now, I'm looking at situations involving etal) it needs to be italicized as a foreign language phrase, and also separated by the last named author with a semicolon. This is not happening currently. &mdash; Ipoellet (talk) 07:19, 2 September 2015 (UTC)


 * Commonly used in English so not italicized. See Manual of Style/Text formatting.


 * The separator issue has been addressed in Module:Citation/CS1/sandbox. See Help talk:Citation Style 1/Archive 8


 * —Trappist the monk (talk) 11:21, 2 September 2015 (UTC)
 * With regard to the separator issue, you'll have to forgive me for not being conversant in developer in-speak. No separator presently appears when using the template. Are you saying that a fix is in progress and is waiting to be migrated? If not, I need a translation. &mdash; Ipoellet (talk) 18:38, 2 September 2015 (UTC)
 * You understood correctly. Modified code sits in the sandbox for a while, typically one to three months, before being moved to the live module. This allows time for debugging and further discussion, and it prevents the module changes from causing every article that uses a CS1 citation template to be refreshed by the WP servers, which is viewed by some as a lot of unnecessary processing. – Jonesey95 (talk) 02:41, 3 September 2015 (UTC)

PLEASE PLEASE PLEASE restore contribution= to journal article citations
For months we have seen complaints here and over on the corresponding CS1 talk pages on 's removal of the ability to use contribution to indicate smaller pieces of journal articles (with title and journal) and for months these have been met with silence and no replacement for this ability. Can we please have some action on this? Without this parameter I am unable to find a valid way to format some references, for instance: (an individual problem by R. Häggkvist, in an article collecting several problems and edited by Mohar, Nowakowski and West, published as a journal paper, and cited as a reference in hypohamiltonian graph). This worked when it was added to the article, but in the current version of the template, this omits several parts of the citation and produces a big red "chapter= ignored" error: There seems to be no way to get both the name of the problem (with its author) and the name of the problem collection (with its editors) formatted as a journal article in the current citation template. Near misses are to pretend that the problem collection is a department of the journal, which at least presents all of the information from the citation but in a bad order that loses the association between which piece of the citation is authored or edited by which people: or to pretend that Discrete Mathematics is really a book series and that the research problems paper in it is really a book, which formats the journal title incorrectly in upright rather than italics, uses book rather than journal formatting for the page numbers, and probably generates completely bogus metadata, but at least presents the information in the correct order. Any other suggestions for how to format this better? For how to persuade Trappist to give us back our functionality? Or for how to get rid of Trappist and find a maintainer for the templates who is more responsive to usability issues? —David Eppstein (talk) 23:26, 27 June 2015 (UTC)
 * What's wrong with you? Do you actually think that you, a subject-area specialist thoroughly familiar with publications in the field and how they are cited, have a better idea of citation formats than does some coder wool-gathering his preconceptions in a vacuum? If he's eliminated the departement parameter then ipso facto you don't need it. Now quit whining and get back to content creation or whatever little people like you do. EEng (talk) 00:02, 28 June 2015 (UTC)
 * Heh. To be fair, the usability and formatting of the bibliography styles required by some of the subject-area publications are even worse. —David Eppstein (talk) 00:13, 28 June 2015 (UTC)
 * Indeed. I really wish we'd stop permitting most of them, and just stick to a consistent citation style, like every other publisher in the world does.  — SMcCandlish ☺ ☏ ¢ ≽ʌⱷ҅ᴥⱷʌ≼  20:51, 30 July 2015 (UTC)
 * This seems like an entirely reasonable fix-it request.  — SMcCandlish ☺ ☏ ¢ ≽ʌⱷ҅ᴥⱷʌ≼  20:52, 30 July 2015 (UTC)


 * So what needs to be done to get this rolling? ~ J. Johnson (JJ) (talk) 20:44, 7 September 2015 (UTC)

Strip use of accessdate param?
Should any use of accessdate by bulk-stripped from page, using AWB?

This seems to be a thing at present ( et al). No reason why that I can see. If an EL fails and has to be recovered from an archive site (a not uncommon event), this date is helpful for recovering the link. Andy Dingley (talk) 14:54, 13 October 2015 (UTC)
 * The stripping appears to be for non-web sources, but instead for journals and books. The documentation says the  field is associated with the   field, and it's a CS1 error to have an accessdate without a url. The main bibliographic reason for accessdate is because a source could change (as well as disappear) and the ref needs to refer to the state if the source at a certain point in time. But hardcopy-published materials are immutable and sources that don't have a url aren't subject to "EL fails". DMacks (talk) 15:02, 13 October 2015 (UTC)
 * Classes of hard copies are not always immutable, nor intact, nor complete—even when one perhaps expects them to be. accessdate gives the confirmation that on that date a retrieved copy indeed contained the claimed information.  Please leave accessdate alone, it's useful when it is present.  —Sladen (talk) 15:24, 13 October 2015 (UTC)

Using URLs in param for cite templates
The fragment:  param for cite templates gives an error. There is work in progress to remove these, by converting the param value to a simple, unlinked, name.

The same thing happens with the  param - although the error message then is confusing, as it still refers to the work param, even though that isn't being explicitly called.

Should the use of an EL in these parameters be permitted and supported? Or at least for

Or, would such a use give two blue links in the result and thus risk confusing readers? Andy Dingley (talk) 16:07, 13 October 2015 (UTC)


 * A conversation on this topic is at HT:CS1


 * —Trappist the monk (talk) 16:46, 13 October 2015 (UTC)

Trans_quote
Will there be a trans_quote field to complement the Trans_title field? --Richard Arthur Norton (1958- ) (talk) 16:10, 30 May 2015 (UTC)
 * You can just use the original parameter: Mi casa es su casa", English: "My house is your house. It's only awkward in that you need to remember that it inserts the beginning and ending  characters for you, which is clearly documented. Maybe that's  awkward, but it doesn't bug me personally, so I'm kind of neutral on it.  — SMcCandlish ☺ ☏ ¢ ≽ʌⱷ҅ᴥⱷʌ≼  20:56, 30 July 2015 (UTC)


 * I support the idea. See this thread for a very similar proposal: Help talk:Citation Style 1/Archive 9
 * --Matthiaspaul (talk) 23:05, 22 October 2015 (UTC)