Template talk:Cite Q/Archive 5

URL override not working when using cite Q for
gives:

in which the title takes us to the wikipedia article and not to the url in the cite Q template.

I think there are some major problems with the wikidata item and would be very pleased to be guided with respect to how to fix it so that cite Q will act in the way I wish. (i.e., direct the reader to the relevant page in the original text) MargaretRDonald (talk) 23:27, 4 January 2021 (UTC)
 * if you add the parameter yes, you can see what is passed from CiteQ to Citation:
 * In this case, the source appears to be a book that has an article on Wikipedia. The usual convention, I think, is to link the title to the article where it exists, and that is how CiteQ behaves. Do you think that it should instead prefer to link to a url parameter when supplied? Should that always be the behaviour for books? So we would get:
 * If that is the preferred behaviour (url takes precedence over Wikipedia article), then it can be coded to do that. Any other thoughts? --RexxS (talk) 01:29, 5 January 2021 (UTC)
 * I'd favor linking to Wikipedia (and Wikisource/Wikibooks) whenever possible. ~ ★ nmaia d 01:53, 5 January 2021 (UTC)
 * If an editor supplies an then shouldn't the  take precedence?  In cs1|2 when given both url and either a linked title or title-link, Module:Citation/CS1 emits an error message and links title with url; title-link, if used, is ignored; linked titles just make a mess:
 * When an editor supplies a url, must override any url from wikidata as well as any value it wants to put in title-link and must not wikilink title.
 * —Trappist the monk (talk) 02:16, 5 January 2021 (UTC)
 * The aim (in my view) is always to link to the external text (& never to an article about the text). MargaretRDonald (talk) 04:36, 5 January 2021 (UTC)
 * I've amended the logic in the sandbox. If I have it right, then when a title is passed as a parameter, it becomes the title displayed. Otherwise, it will display any . Otherwise it will display the local label for the Wikidata item (with language fallback). Otherwise it will display the label in the language of the publication. Otherwise it displays "No label or title -- debug:" for the moment. If we don't see that anywhere, I'll remove it and rely on CS1 to trap any missing parameters. If a url is available, either from a parameter of from Wikidata, the title won't be passed as linked item. I need to work out whether it's best to keep the title unlinked and pass a title-link parameter, which is suspect is the preferred behaviour. Examples:
 * If anybody has any time to test the sandbox with edge case sources additional to Template:Cite Q/testcases, it would be appreciated. --RexxS (talk) 21:41, 5 January 2021 (UTC)
 * Hoping someone will deal with the critical issue that a citation should never link to an article. If it is to link at all, it must be to an external source. In other words, the default link for a title should be to the source at "full work available at" or to the overriding url in the cite Q citation. The behaviour of    remains unsatisfactory. (sorry) MargaretRDonald (talk) 20:58, 8 January 2021 (UTC)
 * Why should a citation never link to an article? Does that proscription appear in one of Wikipedia's policies or guidelines? – Jonesey95 (talk) 21:30, 8 January 2021 (UTC)
 * try WP:CIRC. A citation must not in any way suggest that a Wikipedia article is a reliable source. Peter coxhead (talk) 21:39, 8 January 2021 (UTC)
 * Thanks, I am familiar with WP:CIRC, and with the long-standing title-link that is a feature of the CS1 family of templates and which is used in somewhere over 20,000 articles. The latter is perfectly valid as long as it is used in a way that does not contravene the former. – Jonesey95 (talk) 00:42, 9 January 2021 (UTC)
 * I can see that linking to an article perhaps makes sense. But it makes no sense at all when a book is out of copyright and available at a URL as is the case for many, many  books, available at BHL. Where a book has a "full work available at" property completed, the default behaviour should be to use the url for the title link, and not the wikipedia article. MargaretRDonald (talk) 04:11, 9 January 2021 (UTC)
 * Exactly. It's not the linking to the article that's the problem, it's doing this instead of linking to the url where this exists. Peter coxhead (talk)
 * I'm really not understanding what the problem is. As far as I can see, that is exactly the behaviour I've coded into the sandbox. I'm beginning to think that no matter what fixes I make, it's never going to satisfy you.
 * Please give me an example of where the template isn't doing what you're asking for. --RexxS (talk) 21:54, 9 January 2021 (UTC)
 * yes, the sandbox behaviour is correct. Thanks! Peter coxhead (talk) 07:11, 10 January 2021 (UTC)
 * Thanks, That is exactly the behaviour needed.  (It will be great when migrated from the sandbox.)  Cheers, MargaretRDonald (talk) 19:21, 10 January 2021 (UTC)
 * thank you for the feedback. Now that I know the sandbox is doing the job wanted, I can update the main module from it. Please let me know if new issues arise. --RexxS (talk) 18:23, 11 January 2021 (UTC)
 * There's a CS1 error: invisible character: 'line feed error at position 50 in |id=' on pages using Template:Academic peer reviewed. DrKay (talk) 20:20, 12 January 2021 (UTC)
 * If anybody has any time to test the sandbox with edge case sources additional to Template:Cite Q/testcases, it would be appreciated. --RexxS (talk) 21:41, 5 January 2021 (UTC)
 * Hoping someone will deal with the critical issue that a citation should never link to an article. If it is to link at all, it must be to an external source. In other words, the default link for a title should be to the source at "full work available at" or to the overriding url in the cite Q citation. The behaviour of    remains unsatisfactory. (sorry) MargaretRDonald (talk) 20:58, 8 January 2021 (UTC)
 * Why should a citation never link to an article? Does that proscription appear in one of Wikipedia's policies or guidelines? – Jonesey95 (talk) 21:30, 8 January 2021 (UTC)
 * try WP:CIRC. A citation must not in any way suggest that a Wikipedia article is a reliable source. Peter coxhead (talk) 21:39, 8 January 2021 (UTC)
 * Thanks, I am familiar with WP:CIRC, and with the long-standing title-link that is a feature of the CS1 family of templates and which is used in somewhere over 20,000 articles. The latter is perfectly valid as long as it is used in a way that does not contravene the former. – Jonesey95 (talk) 00:42, 9 January 2021 (UTC)
 * I can see that linking to an article perhaps makes sense. But it makes no sense at all when a book is out of copyright and available at a URL as is the case for many, many  books, available at BHL. Where a book has a "full work available at" property completed, the default behaviour should be to use the url for the title link, and not the wikipedia article. MargaretRDonald (talk) 04:11, 9 January 2021 (UTC)
 * Exactly. It's not the linking to the article that's the problem, it's doing this instead of linking to the url where this exists. Peter coxhead (talk)
 * I'm really not understanding what the problem is. As far as I can see, that is exactly the behaviour I've coded into the sandbox. I'm beginning to think that no matter what fixes I make, it's never going to satisfy you.
 * Please give me an example of where the template isn't doing what you're asking for. --RexxS (talk) 21:54, 9 January 2021 (UTC)
 * yes, the sandbox behaviour is correct. Thanks! Peter coxhead (talk) 07:11, 10 January 2021 (UTC)
 * Thanks, That is exactly the behaviour needed.  (It will be great when migrated from the sandbox.)  Cheers, MargaretRDonald (talk) 19:21, 10 January 2021 (UTC)
 * thank you for the feedback. Now that I know the sandbox is doing the job wanted, I can update the main module from it. Please let me know if new issues arise. --RexxS (talk) 18:23, 11 January 2021 (UTC)
 * There's a CS1 error: invisible character: 'line feed error at position 50 in |id=' on pages using Template:Academic peer reviewed. DrKay (talk) 20:20, 12 January 2021 (UTC)
 * Please give me an example of where the template isn't doing what you're asking for. --RexxS (talk) 21:54, 9 January 2021 (UTC)
 * yes, the sandbox behaviour is correct. Thanks! Peter coxhead (talk) 07:11, 10 January 2021 (UTC)
 * Thanks, That is exactly the behaviour needed.  (It will be great when migrated from the sandbox.)  Cheers, MargaretRDonald (talk) 19:21, 10 January 2021 (UTC)
 * thank you for the feedback. Now that I know the sandbox is doing the job wanted, I can update the main module from it. Please let me know if new issues arise. --RexxS (talk) 18:23, 11 January 2021 (UTC)
 * There's a CS1 error: invisible character: 'line feed error at position 50 in |id=' on pages using Template:Academic peer reviewed. DrKay (talk) 20:20, 12 January 2021 (UTC)

, are you posting this on the correct talk page? It would help a lot if you would link to an article displaying this error. – Jonesey95 (talk) 20:35, 12 January 2021 (UTC)
 * I'm talking about the change to Module:Cite Q, it caused an error on 121 articles transcluding Template:Academic peer reviewed. DrKay (talk) 20:49, 12 January 2021 (UTC)
 * Returning to the issue of pointing to articles rather than sources... Yesterday this aspect of Cite Q was working perfectly, but now we are back to pointing to articles. See Westringia rigida. Hoping the aspects of so appreciated will return soon (including "stated as" for publisher) MargaretRDonald (talk) 20:26, 14 January 2021 (UTC)
 * DrKay reverted my update of the main module from the sandbox. I'm not going to edit-war over it. --RexxS (talk) 23:56, 14 January 2021 (UTC)
 * One is citing a journal paper or a book and NOT a page in wikipedia. As I have repeatedly said, my view is that a citation should not point to a wikipedia page. And again, as stated above, there are many hundreds if not thousands of books which are available online and to which a citation may point. When an editor places a url in a citeQ template then citeQ should always point to that url and in my view when wikidata includes the "full work available at" property this too should also override any pointing to a wikipedia page. If this cannot be fixed I will be giving up on cite Q for books which is a pity. (Thanks for all your work on this...) MargaretRDonald (talk) 21:30, 21 January 2021 (UTC)
 * You must be looking at cached pages. Try purging. DrKay (talk) 22:08, 21 January 2021 (UTC)
 * I'm sorry, but I refuse to try to track down any more complaints that don't quote a concrete example of the behaviour, which I need so that I can figure out where the problem is. I'm sorry to lose your enthusiastic support of CiteQ, but I've already wasted a week on tracking down a problem that wasn't even in the template I was working on, and I don't intend to carry on chasing hypotheticals. --RexxS (talk) 22:18, 21 January 2021 (UTC)
 * My most humble apologies. (I was responding to "DrKay reverted my update of the main module from the sandbox. I'm not going to edit-war over it"). Things are working perfectly: see Westringia rigida, and I am also delighted to be able to now be able to italicise within titles. MargaretRDonald (talk) 22:51, 21 January 2021 (UTC)

[Update:] It turns out that Template:Cite Q EtAl, which uses a vertical layout for its parameters, was adding an extra line feed and space onto the positional parameter (which eventually becomes  ). The old version of CiteQ wasn't sensitive to that, but the new version in the sandbox was. I've now fixed Cite Q EtAl, so I'm going to try updating CiteQ from its sandbox again. Please report any issues. Thanks. --RexxS (talk) 21:05, 21 January 2021 (UTC)
 * Thank you for tracking that down. DrKay (talk) 21:25, 21 January 2021 (UTC)

Cite Q enabling easy translation
I am grateful to User:Tetizeraz who tells me that they have translated Richard Henry Yapp to Portuguese. As far as I know, it's the first time that an article with all of its citations in Cite Q has been translated from one language to another. Andy Mabbett ( Pigsonthewing ); Talk to Andy; Andy's edits 15:09, 17 January 2021 (UTC)
 * Yup. I could copy paste Cite Q translations just like normal citations. The only minor problem in Portuguese Wikipedia is that I had to change all instances of 'Cite Q' to 'Citar Q', but I believe this will be solved in the future. For now, there's the workaround of changing all instances of a phrase to another in the wiki editor. User:Tetizeraz. Send me a ✉️ ! 16:03, 17 January 2021 (UTC)
 * you might want to see if there's a fix for the 'problem' mentioned above. User:Tetizeraz. Send me a ✉️ ! 16:04, 17 January 2021 (UTC)
 * pt:Predefinição:Cite Q exists as a redirect, so you should not need to do that. Andy Mabbett ( Pigsonthewing ); Talk to Andy; Andy's edits 19:19, 20 January 2021 (UTC)
 * If, for example, an English language work (lets say a translation of a work by Pope Gregory XIII) is cited in the French Wikipedia using Cite Q, will the author's name be written in English or French (Grégoire XIII)? It seems to me the name should match the way it is written in the work being cited so people verifying material will be certain they are looking at the correct work. Jc3s5h (talk) 16:55, 17 January 2021 (UTC)
 * The module normally draws information from Wikidata using the authors' labels for, so are language dependent, although if a name is provided by , it is a language-independent string which overrides the label. If I look at using Cite Q/sandbox on my Commons sandbox with my language set to French, I get  . If the person is verifying the citation on the French Wikipedia, I would expect them to understand that the author is the person they recognise as Grégoire XIII, so I don't see much of a problem. Otherwise, the fix is to add the qualifier  to the  statement for the citation on Wikidata. --RexxS (talk) 23:45, 17 January 2021 (UTC)
 * The module normally draws information from Wikidata using the authors' labels for, so are language dependent, although if a name is provided by , it is a language-independent string which overrides the label. If I look at using Cite Q/sandbox on my Commons sandbox with my language set to French, I get  . If the person is verifying the citation on the French Wikipedia, I would expect them to understand that the author is the person they recognise as Grégoire XIII, so I don't see much of a problem. Otherwise, the fix is to add the qualifier  to the  statement for the citation on Wikidata. --RexxS (talk) 23:45, 17 January 2021 (UTC)

, I think Cite Q will help for easy translation in the future between certain wikis to English Wikipedia. I just created and translated an article for English Wikipedia and it's a mess. See diff. I asked the Teahouse to see if they can fix this.User:Tetizeraz. Send me a ✉️ ! 12:51, 23 January 2021 (UTC)

Deployment in other wikipedias
I am hoping that will be deployed on viwiki warwiki svwiki and cebwiki. Is there any chance of this happening? MargaretRDonald (talk) 00:44, 29 December 2020 (UTC)
 * I don't know those wikis well, and although I have some experience in deploying WikidataIB on some other language wikis, I'm not an expert in languages. My advice would be to find somebody who is native to each wiki who is willing to take on the job of deployment. It might be worth pinging, a native Brazilian speaker with excellent English, who has implemented CiteQ and WikidataIB on the ptwiki, for some ideas of the possible stumbling blocks. --RexxS (talk) 18:25, 31 December 2020 (UTC)
 * Hello! I would be super happy to get help to update the version we have on svwiki. We have a version imported from March 2020, but practically no use in main namespace, so we could just replace it with the latest version. Ainali (talk) 15:29, 24 January 2021 (UTC)
 * I'm happy to help as long as you don't need me to learn Swedish. You could try importing the module version dated 2021-01-20 into your sandbox and test it out. We've tried hard to internationalise as much as we can, but I'd love to get feedback from any problems with transwiki deployment. --RexxS (talk) 19:20, 24 January 2021 (UTC)
 * I can handle the Swedish. But I am not sure which module you are referring to, Module:Cite Q seems to have page revisions from the January 12 and 21, but not 20. Ainali (talk)
 * Sorry, I should have made it clearer: I add a version date at the top of these modules and I usually test changes in the sandbox first (Module:Cite Q/sandbox). It means that the main module may actually be updated a day or two later, but the date stamp I add at the top of the module will correspond to the sandbox date (so that they are clearly identifiable as the same version). The latest version is "-- Version: 2021-01-20". --RexxS (talk) 21:48, 24 January 2021 (UTC)
 * Thanks, I copied it to sv:Modul:Cite Q/Sandlåda. Ainali (talk)
 * I tried it in sv:Användare:Ainali/sandbox but get an error: "Luafel i Modul:Cite_Q/Sandlåda på rad 360: attempt to call upvalue 'getPropOfProp' (a nil value).". Ainali (talk) 21:59, 24 January 2021 (UTC)
 * Unfortunately, CiteQ depends on Module:WikidataIB and you have an old version from March 2020, so I've made a short fix to it to get 'getPropOfProp' working. However, that in turn relies on Module:Complex date to internationalise dates. At this point, I really need to ping to see if he can help you get a working version of Complex date on the Swedish Wikipedia. --RexxS (talk) 22:39, 24 January 2021 (UTC)
 * I can move Module:Complex date to Swedish Wikipedia. I am also working on c:Module:Cite Wikidata which is a minimalist alternative to CiteQ with all the internationalization done through c:Data:I18n/Cite Wikidata.tab. I can port that one to Swedish wiki as well. --Jarekt (talk) 01:39, 25 January 2021 (UTC)
 * Thanks, I copied it to sv:Modul:Cite Q/Sandlåda. Ainali (talk)
 * I tried it in sv:Användare:Ainali/sandbox but get an error: "Luafel i Modul:Cite_Q/Sandlåda på rad 360: attempt to call upvalue 'getPropOfProp' (a nil value).". Ainali (talk) 21:59, 24 January 2021 (UTC)
 * Unfortunately, CiteQ depends on Module:WikidataIB and you have an old version from March 2020, so I've made a short fix to it to get 'getPropOfProp' working. However, that in turn relies on Module:Complex date to internationalise dates. At this point, I really need to ping to see if he can help you get a working version of Complex date on the Swedish Wikipedia. --RexxS (talk) 22:39, 24 January 2021 (UTC)
 * I can move Module:Complex date to Swedish Wikipedia. I am also working on c:Module:Cite Wikidata which is a minimalist alternative to CiteQ with all the internationalization done through c:Data:I18n/Cite Wikidata.tab. I can port that one to Swedish wiki as well. --Jarekt (talk) 01:39, 25 January 2021 (UTC)

ASIN
This template should suppress the ASIN when the ISBN is present. AManWithNoPlan (talk) 15:36, 20 January 2021 (UTC)
 * thank you for spotting that. I've coded the change in the sandbox, ready for testing. --RexxS (talk) 16:46, 20 January 2021 (UTC)
 * Why should it do this? Unless someone erroneously put an ISBN value into the ASIN parameter (something that should be corrected in WD then, but is also treated accordingly inside citation), they are two different identifiers and should be treated as such, that is, retrieved from WD and passed down to citation. If someone does not want a particular value to be reported in a citation, s/he should use unset, unset, etc. to mute that value.
 * --Matthiaspaul (talk) 04:11, 25 January 2021 (UTC)
 * We would do this because editors have a preference for using an identifier linking to a free resource (isbn links to Special:BookSources) where possible, rather than a link to a commercial entity such as Amazon. I can't think of an example where both an isbn and an asin would be useful in the same citation. --RexxS (talk) 16:27, 25 January 2021 (UTC)
 * ASINs can be looked up for free - as can ISBNs. The corresponding publications, however, are not free in both cases. The assignment of ISBNs costs money, therefore free publications do not normally carry ISBNs (at least not until decades after their publication).
 * I see your point of not unnecessarily routing readers to a particular distributor if a generic resource is available at the same time. But I also think that it should not be our business to algorithmically encode our personal preferences in a template automatically pulling data from WD instead of leaving it up to the editorial judgement of the article editors if some information should be suppressed or not. In our role here we are template developers, not article editors, and to maintain a neutral point of view, we should just report all usable information available at WD, instead of filtering out some information without being asked to do so by article editors through template parameters (like ). If an ASIN entry does not make sense in a particular scenario, it should not be listed at WD in the first place. So, if an ASIN entry exists at WD, some editor probably had a reason to add it. citation does not mute asin if isbn is given, nor do we do this for other identifiers, so we shouldn't do this, either.
 * --Matthiaspaul (talk) 18:15, 25 January 2021 (UTC)
 * It's good that you can see the point of not linking to a commercial retailer like Amazon when they are only one of several options, including World Cat, especially as library membership tends to be free in the English-speaking world.
 * Its not my personal preference. It's a general preference expressed by editors many times over the years. It's going to need an explicit statement of consensus to convince me that we should be linking to both isbn and asin. Here's what Help:Citation Style 1 says about its use of the asin parameter: "if first character of asin value is a digit, use isbn. Because this link favours one specific distributor, include it only if standard identifiers are not available."
 * I'm an editor, not a template developer, and that's my role as a volunteer on the English Wikipedia.
 * I don't agree that there is an NPOV argument in favour of including a single retailer's product identification when more neutral identifiers are available. Previous discussions:
 * Templates for discussion/Log/2012 June 25
 * Help talk:Citation Style 1/Archive 6
 * Help talk:Citation Style 1/Archive 44
 * I don't see anything there that indicates that asin is to be used other than in the absence of other, more general identifiers. YMMV. --RexxS (talk) 20:59, 25 January 2021 (UTC)
 * And the most recent Templates for discussion/Log/2017 April 13 and Wikipedia talk:WikiProject Spam/2017 Archive May 1. &#32; Headbomb {t · c · p · b} 21:09, 25 January 2021 (UTC)
 * Rexx, people often have multiple roles, sometimes even (potentially) conflicting ones. That's all fine for as long as we can keep them separate depending on context. In this edit you acted in your (voluntary) role as template developer, not as article editor, but applied code to the template to preempt a decision that should be made in the domain of an article, among article editors (including you in your role as article editor), and through other means such as adding unset if that's what the article editors decide on.
 * The point I was trying to make here (that is, on the cite Q talk page, in my role as (another) template editor) is not if or if not ASINs are appropriate in a citation in an article. Sure, I have a opinion on this as well, and as article editor it might not be that different from yours, but in our role as template developers we should be neutral in regard to what information to include in a citation, and therefore pass through what's available. We shouldn't apply filters, even if, in our role as article editors, muting a particular value might seem desirable. Let's leave it to those working on an article to decide this on an individual case-by-case basis. --Matthiaspaul (talk) 23:37, 25 January 2021 (UTC)
 * Editors can decide to include it on an individual case-by-case basis, by setting it manually within the template. However, it makes to default to the general preference, as expressed by RexxS above, to omit when ISBN is available. Passing through every possible available value, even when it is known that certain ones will generally not be wanted, would not promote "neutrality" so much as it would promote non-use of this template. Nikkimaria (talk) 00:31, 26 January 2021 (UTC)
 * Matthias, I am an editor on the English Wikipedia, with no special privilege as an editor. There is no such role as "template developer", nor should editors attempt to give themselves fancy titles as if they are any different from any other editor. Anybody who disagrees with my edit can change it or debate it, and then we continue the process of finding consensus for what the content should be. That is the same process as any other piece of content on Wikipedia in any namespace, and so it should be.
 * You are too invested in your own scheme for writing  to see the bigger picture: nobody except you has suggested that there is any point in having both isbn and asin; and everybody except you agree that where both exist, we should simply use isbn. You have been given links to dozens of editors' views on linking to Amazon, and I could not find a single person in those debates who suggested using asin where other more general identifiers exist. That is as clear a consensus as I can imagine, and I don't see my edit as representing anything other than the near unanimous will of those editors who have considered the issue. So please, if you somehow think that there is any value in a citation containing both isbn and asin, give me an example of such a citation and explain to me why that example is improved by the presence of the direct link to Amazon.
 * Before you continue with your argument that "we should make the choice available so that editors on an article can decide for themselves", consider first that most editors are not familiar with  and that every time we have to explicitly opt-out of fetching a value from Wikidata, we give yet another piece of ammunition to the considerable body of editors who want nothing to do with Wikidata and nearly succeeded in having all Wikidata sourced content banned from the English Wikipedia not so long ago. --RexxS (talk) 03:22, 26 January 2021 (UTC)
 * You are too invested in your own scheme for writing  to see the bigger picture: nobody except you has suggested that there is any point in having both isbn and asin; and everybody except you agree that where both exist, we should simply use isbn. You have been given links to dozens of editors' views on linking to Amazon, and I could not find a single person in those debates who suggested using asin where other more general identifiers exist. That is as clear a consensus as I can imagine, and I don't see my edit as representing anything other than the near unanimous will of those editors who have considered the issue. So please, if you somehow think that there is any value in a citation containing both isbn and asin, give me an example of such a citation and explain to me why that example is improved by the presence of the direct link to Amazon.
 * Before you continue with your argument that "we should make the choice available so that editors on an article can decide for themselves", consider first that most editors are not familiar with  and that every time we have to explicitly opt-out of fetching a value from Wikidata, we give yet another piece of ammunition to the considerable body of editors who want nothing to do with Wikidata and nearly succeeded in having all Wikidata sourced content banned from the English Wikipedia not so long ago. --RexxS (talk) 03:22, 26 January 2021 (UTC)

Can you enable some additional field that lists a corrigendum or retraction?
Perhaps the reference ends with a suggestion that the work might have been affected by the corrigendum or the retraction. I've spent a lot of time focused on retractions and errata in Wikidata, but the impact on cited material in Wikipedia may be trivial. Example:

(see also Q98732655)

Trilotat (talk) 21:23, 21 January 2021 (UTC)
 * I've knocked up a demo in the sandbox to explore your suggestion:
 * I don't want to make the citation longer if I can help it, so I've used a piped link to the word "(erratum)". At present, it's linking to the Wikidata entry for the erratum, which is less than ideal. Perhaps we should link to the erratum's doi, assuming that every erratum will have a doi? Would we miss out many if we only created the link where the erratum's doi exists? --RexxS (talk) 22:53, 21 January 2021 (UTC)
 * I don't want to make the citation longer if I can help it, so I've used a piped link to the word "(erratum)". At present, it's linking to the Wikidata entry for the erratum, which is less than ideal. Perhaps we should link to the erratum's doi, assuming that every erratum will have a doi? Would we miss out many if we only created the link where the erratum's doi exists? --RexxS (talk) 22:53, 21 January 2021 (UTC)
 * I don't want to make the citation longer if I can help it, so I've used a piped link to the word "(erratum)". At present, it's linking to the Wikidata entry for the erratum, which is less than ideal. Perhaps we should link to the erratum's doi, assuming that every erratum will have a doi? Would we miss out many if we only created the link where the erratum's doi exists? --RexxS (talk) 22:53, 21 January 2021 (UTC)
 * I don't want to make the citation longer if I can help it, so I've used a piped link to the word "(erratum)". At present, it's linking to the Wikidata entry for the erratum, which is less than ideal. Perhaps we should link to the erratum's doi, assuming that every erratum will have a doi? Would we miss out many if we only created the link where the erratum's doi exists? --RexxS (talk) 22:53, 21 January 2021 (UTC)


 * My vote is that I think it would be fine to only create the link where the erratum's doi exists. I've always had this creeping suspicion that an unknown number of cited references have been revoked or amended, but there's nothing to that point in wikipedia. Clearly, if you read a paper article, you'll never know if it's been revoked, but I think Wikipedia has the technology opportunity to make us aware if the source has been corrected or retracted. Thanks for your time. Trilotat (talk) 23:39, 21 January 2021 (UTC)
 * If I understand correctly, doi only applies to journal articles, not books. Books do have published errata. For example, the errata for are located at http://www.uscibooks.com/seiderr.htm Jc3s5h (talk) 00:42, 22 January 2021 (UTC)
 * For what it's worth, points to errata for a published book, the DOI of which is . – Jonesey95 (talk) 00:55, 22 January 2021 (UTC)
 * although I think I'd prefer to link to a document object identifier where it exists, I'm still unclear over what the preferred behaviour should be where it doesn't. Do you want to see a link to a url or to the erratum's Wikidata entry? If you look at, there's no mention of an erratum, so I don't know how we would be expected to read its location for use in CiteQ.
 * you said "I've always had this creeping suspicion that an unknown number of cited references have been revoked or amended, but there's nothing to that point in wikipedia." Did you know that if a paper is retracted or replaced and that is noted in Wikidata (properties: and ), then CiteQ will automatically place the article where the citation appears in the category Category:Cite Q - cites a retracted work or Category:Cite Q - cites a replaced work as appropriate? --RexxS (talk) 19:53, 23 January 2021 (UTC)
 * I notice that Citation does not make any provision for errata. So today, in an article that uses Citation, one could write two separate citations, one for the main work, and another for the errata. Or, one could add information about the errata in the reference, but outside the Citation template (probably right after the template). So I think the place to begin would be Citation, and once that is the way we want it, Cite Q can do the same thing. Jc3s5h (talk) 20:02, 23 January 2021 (UTC)
 * That means we would be waiting on changes to another template to implement a request here. Generally, I'm not in favour of that when we have the means to implement the request ourselves. As you say, the obvious work-around is simply to write, but I think  was looking for CiteQ to make use of  where someone has taken the trouble of adding it to Wikidata. --RexxS (talk) 20:16, 23 January 2021 (UTC)
 * Yes, that's what I'm hoping for. A number of us Wikidata editors are making an effort to connect items to their retractions and corrections, so Cite Q will benefit if/when you make this happen. Thanks for considering my request. Trilotat (talk) 20:40, 23 January 2021 (UTC)
 * The DOI appendage "_6" appears to be Springer's notation for chapter DOIs only, not some general notation to refer to errata (that is, Chapter 6 refers to "Errata" in this particular example only).
 * So far, I added errata info immediately following the citation in the form "Errata: ". For a consistent format, we should probably add some kind of errata-url/errata-doi parameter to citation. Until this happens, cite Q could use a three-stage model for the link by linking to the errata URL if given at WD, else the errata DOI, or falling back to the errata entry at WD itself.
 * --Matthiaspaul (talk) 04:48, 25 January 2021 (UTC)
 * , you mentioned that items might end up listed in Category:Cite Q - cites a retracted work or Category:Cite Q - cites a replaced work as appropriate, but is there way to list items in a category that have Wikidata property ?
 * Sure, . I've added Category:Cite Q - cites a work with an erratum to the sandbox code, ready for testing. This page already shows up. --RexxS (talk) 00:31, 30 January 2021 (UTC)
 * , I have added a reference to Carapa procera that has been corrected (with a change that has no impact on the reference.) I checked, but it is not in the category. Trilotat (talk) 00:47, 30 January 2021 (UTC)
 * as the code is only in the sandbox for now, you have to test by adding the reference with Cite Q/sandbox to see the effects of the code. I've just amended Carapa procera to use the sandbox, and it's showing up in Category:Cite Q - cites a work with an erratum. Cheers --RexxS (talk) 01:08, 30 January 2021 (UTC)

How to make "published in" work
If I'm trying to Cite Q reference an article, such as:


 * (which is cited by Walker Lane)

How do I ensure that it includes this edited volume info?



I'm not sure if I should put the volume in the published in field in Wikidata. Thanks, Trilotat (talk) 22:56, 29 January 2021 (UTC)
 * At present, CiteQ looks for information in whatever is in for the following properties:, , , , , , , . But if you look at , you find that  points to , which contains no useful information. I would suggest changing that to point to  if you want to make use of the information in there. The property  isn't particularly useful for an article published in a larger collection, as it's a rather general property, particularly when the reciprocal property  isn't present in the target.
 * Would you like to try that as a first step and see if it does what you wanted? If not, I'll look for another way of getting the information you want. --RexxS (talk) 23:46, 29 January 2021 (UTC)
 * Shouldn't Q60826200 also include an isbn? The two  templates render as ; not right for this source.
 * —Trappist the monk (talk) 23:51, 29 January 2021 (UTC)
 * I have updated the items in Wikidata, so it appears to be partly resolved. I don't know what to do to deal with the template rendering issue that Trappist the monk notes. Thank you.Trilotat (talk) 00:26, 30 January 2021 (UTC)
 * The code is very rudimentary, but it guesses that the template should be cite journal only if the Wikidata entry contains . If there's an ISBN, that takes precedence and switches to cite book. If I can find other hints for the template type, I'd be happy to incorporate them. The template choice can also be set by using the template in CiteQ:
 * Obviously not a magazine, but that's just to demo the functionality. --RexxS (talk) 01:01, 30 January 2021 (UTC)
 * In the first example, creates a  (as it should) but then includes these:
 * Both are wrong.
 * is not a journal so the template should not include journal (or any other 'periodical'-like parameter). Because  uses journal for   and because title is  title gets  .  This misassignment shows up in the citation's metadata where the chapter title is assigned to the book title's   key:
 * But, the metadata's  key is omitted because there isn't a chapter (or alias).
 * What should happen for this chapter cite is:
 * and
 * No doubt, Q95597598 at wikidata is an instance of an 'academic chapter' that is 'part of' Q60826200. It is not 'published in' GSA Field Guide because that is a series of books (that value belongs, if it is used, in serial).
 * Yeah, it's a tough nut to crack; if it weren't, Wikicite would already be up and running.
 * —Trappist the monk (talk) 16:59, 30 January 2021 (UTC)
 * In the first example, creates a  (as it should) but then includes these:
 * Both are wrong.
 * is not a journal so the template should not include journal (or any other 'periodical'-like parameter). Because  uses journal for   and because title is  title gets  .  This misassignment shows up in the citation's metadata where the chapter title is assigned to the book title's   key:
 * But, the metadata's  key is omitted because there isn't a chapter (or alias).
 * What should happen for this chapter cite is:
 * and
 * No doubt, Q95597598 at wikidata is an instance of an 'academic chapter' that is 'part of' Q60826200. It is not 'published in' GSA Field Guide because that is a series of books (that value belongs, if it is used, in serial).
 * Yeah, it's a tough nut to crack; if it weren't, Wikicite would already be up and running.
 * —Trappist the monk (talk) 16:59, 30 January 2021 (UTC)
 * and
 * No doubt, Q95597598 at wikidata is an instance of an 'academic chapter' that is 'part of' Q60826200. It is not 'published in' GSA Field Guide because that is a series of books (that value belongs, if it is used, in serial).
 * Yeah, it's a tough nut to crack; if it weren't, Wikicite would already be up and running.
 * —Trappist the monk (talk) 16:59, 30 January 2021 (UTC)
 * Yeah, it's a tough nut to crack; if it weren't, Wikicite would already be up and running.
 * —Trappist the monk (talk) 16:59, 30 January 2021 (UTC)
 * —Trappist the monk (talk) 16:59, 30 January 2021 (UTC)

Cite Q not mentioning the language of the link
In FINLAY-FR-2, I use one Cite Q for one link, it's the third ref. I added, on wikidata, that it is written in Brazilian Portuguese and even the Label only exists in Portuguese and Brazilian Portuguese. So why Cite Q doesn't mention that? User:Tetizeraz. Send me a ✉️ ! 16:13, 29 January 2021 (UTC)
 * the problem was that Wikidata might store the language code in either in or  and CiteQ was only looking for the former. I've updated the code in the sandbox ready for testing:
 * Perhaps you could check to see if works for you in other articles? --RexxS (talk) 20:20, 29 January 2021 (UTC)
 * See here, it doesn't mention the language in other places. You might want to implement your fix in the sandbox. User:Tetizeraz. Send me a ✉️ ! 00:19, 30 January 2021 (UTC)
 * I added the sandbox version of CiteQ to your sandbox (hope you don't mind). It shows the language. I presently have a number of changes pending in the sandbox, and I'd like to make sure they are all well tested before I put them into the main module. Any help in testing would be much appreciated. --RexxS (talk) 00:46, 30 January 2021 (UTC)
 * See diff here. It still doesn't mention Portuguese as a the language of the websites linked. User:Tetizeraz. Send me a ✉️ ! 01:12, 30 January 2021 (UTC)
 * as the code is only in the sandbox for now, you have to test by adding the reference with Cite Q/sandbox to see the effects of the code.
 * The first one has the property set to  and so CiteQ finds the language. The second one doesn't have the property set and so CiteQ doesn't find it. --RexxS (talk) 01:23, 30 January 2021 (UTC)
 * I fixed, it should show the language of the work now. User:Tetizeraz. Send me a ✉️ ! 01:26, 30 January 2021 (UTC)
 * I'm working on finding the language from the language of the title (as it's monolingual text), but that will have be a job for tomorrow. So far I have:
 * I'll have to figure out what to do if there are more than one title in more than one language. --RexxS (talk) 02:08, 30 January 2021 (UTC)
 * You mean Label or ? I can add Portuguese as a title temporarily in those QIDs I created just to test them. Please ping me when you're online so we can do this fast. User:Tetizeraz. Send me a ✉️ ! 15:58, 30 January 2021 (UTC)
 * No, there's no good way of finding the language of a work from its Wikidata label(s). I mean the property, which may or may not be present, but when it is, it has a language associated with it. I'm not yet ready to try that out yet, so you shouldn't worry about adding things to Wikidata to improve language recognisability other than  where appropriate. --RexxS (talk) 18:08, 30 January 2021 (UTC)
 * Sure, I'll wait and see then. User:Tetizeraz. Send me a ✉️ ! 18:11, 30 January 2021 (UTC)
 * I'll have to figure out what to do if there are more than one title in more than one language. --RexxS (talk) 02:08, 30 January 2021 (UTC)
 * You mean Label or ? I can add Portuguese as a title temporarily in those QIDs I created just to test them. Please ping me when you're online so we can do this fast. User:Tetizeraz. Send me a ✉️ ! 15:58, 30 January 2021 (UTC)
 * No, there's no good way of finding the language of a work from its Wikidata label(s). I mean the property, which may or may not be present, but when it is, it has a language associated with it. I'm not yet ready to try that out yet, so you shouldn't worry about adding things to Wikidata to improve language recognisability other than  where appropriate. --RexxS (talk) 18:08, 30 January 2021 (UTC)
 * Sure, I'll wait and see then. User:Tetizeraz. Send me a ✉️ ! 18:11, 30 January 2021 (UTC)
 * Sure, I'll wait and see then. User:Tetizeraz. Send me a ✉️ ! 18:11, 30 January 2021 (UTC)

Bug: multiple ISBNs
I think multiple ISBNs are sent as a comma-delimited list, and the comma throws an error “Check &#124;isbn= value: invalid character (help).”



Yes, I know an edition should have one ISBN. But there are entries for multi-volume works, with different ISBNs for the set and each volume, and there are enough that are virtually identical that this condition will remain perennial over there. So it probably shouldn’t throw an ugly error over here, where most readers and editors can’t fix it, and the “help” link doesn’t help. —Michael Z. 00:13, 31 January 2021 (UTC)


 * It only happens with ISBN-10, and I could change the separator to space, but this is what you would get:
 * So citation only deals with one ISBN. I've amended the CiteQ sandbox to only take the first best value:
 * Nevertheless, you can always override the Wikidata value with a manual parameter:
 * There's not much else I can do. Technically, Wikidata should maintain separate entries for hardback and paperback editions when each has a different ISBN. --RexxS (talk) 01:43, 31 January 2021 (UTC)
 * Nevertheless, you can always override the Wikidata value with a manual parameter:
 * There's not much else I can do. Technically, Wikidata should maintain separate entries for hardback and paperback editions when each has a different ISBN. --RexxS (talk) 01:43, 31 January 2021 (UTC)
 * There's not much else I can do. Technically, Wikidata should maintain separate entries for hardback and paperback editions when each has a different ISBN. --RexxS (talk) 01:43, 31 January 2021 (UTC)
 * There's not much else I can do. Technically, Wikidata should maintain separate entries for hardback and paperback editions when each has a different ISBN. --RexxS (talk) 01:43, 31 January 2021 (UTC)
 * There's not much else I can do. Technically, Wikidata should maintain separate entries for hardback and paperback editions when each has a different ISBN. --RexxS (talk) 01:43, 31 January 2021 (UTC)


 * Thank you. That looks like a fine solution. For myself, I don’t want to see multiple ISBNs in a citation (or indeed any, since I can click through to Wikidata). —Michael Z. 02:32, 31 January 2021 (UTC)

Trim space before/after parameter
I noticed above that a space before or after the parameter results in a space before/after the Qid in the output of the template. Can this be trimmed please? &mdash; Martin (MSGJ · talk) 22:56, 1 February 2021 (UTC)


 * The sandbox should cope with that now:
 * Although there's something odd about how the cs1 style cite book in the second example appears to have an extra space at the end. I'll investigate tomorrow. --RexxS (talk) 03:22, 2 February 2021 (UTC)
 * cs1 ( etc) terminates rendered citations with a dot; cs2 does not render a terminating character.  Second example has an ISBN so uses.
 * Note: when the underlying template is cs1, cs1 is superfluous. Same applies to cs2.  It hasn't happened yet but I expect that sometime in future, someone will add code to cs1|2 that will detect and flag superfluous mode parameters.
 * —Trappist the monk (talk) 12:05, 2 February 2021 (UTC)
 * thanks for the point about mode. At present, the code is very poor at detecting which cite template to use, so I wanted to make sure that editors had the opportunity to set the mode appropriately to match a given article, and that whatever was set would take precedence over what the code thought the variant of cs1 should be. If we can improve the variant prediction significantly, perhaps we will be able to trust that and drop the redundant mode parameters passed. --RexxS (talk) 14:47, 2 February 2021 (UTC)
 * Note: when the underlying template is cs1, cs1 is superfluous. Same applies to cs2.  It hasn't happened yet but I expect that sometime in future, someone will add code to cs1|2 that will detect and flag superfluous mode parameters.
 * —Trappist the monk (talk) 12:05, 2 February 2021 (UTC)
 * thanks for the point about mode. At present, the code is very poor at detecting which cite template to use, so I wanted to make sure that editors had the opportunity to set the mode appropriately to match a given article, and that whatever was set would take precedence over what the code thought the variant of cs1 should be. If we can improve the variant prediction significantly, perhaps we will be able to trust that and drop the redundant mode parameters passed. --RexxS (talk) 14:47, 2 February 2021 (UTC)
 * thanks for the point about mode. At present, the code is very poor at detecting which cite template to use, so I wanted to make sure that editors had the opportunity to set the mode appropriately to match a given article, and that whatever was set would take precedence over what the code thought the variant of cs1 should be. If we can improve the variant prediction significantly, perhaps we will be able to trust that and drop the redundant mode parameters passed. --RexxS (talk) 14:47, 2 February 2021 (UTC)

Request for Vancouver support
I have reservations about using this template, but if it is going to be widely deployed, per WP:CITEVAR, adding support for the pass-through parameter vanc is essential. Boghog (talk) 09:50, 6 February 2021 (UTC)
 * It supports passing through any of the parameters that the CS1/2 family support.
 * This is what is passed to Cite book:
 * CITEVAR already makes it impossible for the template to be widely deployed at present. --RexxS (talk) 10:05, 6 February 2021 (UTC)
 * This is what is passed to Cite book:
 * CITEVAR already makes it impossible for the template to be widely deployed at present. --RexxS (talk) 10:05, 6 February 2021 (UTC)
 * This is what is passed to Cite book:
 * CITEVAR already makes it impossible for the template to be widely deployed at present. --RexxS (talk) 10:05, 6 February 2021 (UTC)
 * CITEVAR already makes it impossible for the template to be widely deployed at present. --RexxS (talk) 10:05, 6 February 2021 (UTC)
 * CITEVAR already makes it impossible for the template to be widely deployed at present. --RexxS (talk) 10:05, 6 February 2021 (UTC)

Thanks, I now see vanc does work, but only if no spaces are included in the template. When I first tested it, I padded the parameter with spaces, and it did not work. Boghog (talk) 10:50, 6 February 2021 (UTC)
 * For, all that vanc does is change the author separator from semicolon to comma. Until wikidata provides separate surname / given name data, it will never really support Vancouver style:
 * Frankel KL, Glazner AF, Kirby E, ...
 * White space in parameter values is ignored because all cs1|2 (and so all ) parameters are named. Mediawiki trims leading and trailing white space from named parameters before they are delivered to templates/modules.  Your example template renders exactly the same regardless of whitespace.
 * —Trappist the monk (talk) 11:38, 6 February 2021 (UTC)

I now see the problem was that I used the deprecated parameter name (name-list-format, should be name-list-style). The problem was not with spaces. Sorry for the confusion. Boghog (talk) 11:46, 6 February 2021 (UTC)
 * No. name-list-format is not (yet) deprecated though it will be at the next cs1|2 module suite update.  Right now both of name-list-format and name-list-style are aliases and both work (incompletely as I noted above).  We can test this:
 * Both of those renderings are the same, and both of them are still not proper Vancouver style citations.
 * —Trappist the monk (talk) 12:00, 6 February 2021 (UTC)
 * Both of those renderings are the same, and both of them are still not proper Vancouver style citations.
 * —Trappist the monk (talk) 12:00, 6 February 2021 (UTC)
 * Both of those renderings are the same, and both of them are still not proper Vancouver style citations.
 * —Trappist the monk (talk) 12:00, 6 February 2021 (UTC)

Use P1343 as reference
(Cross-posted at Module talk:Wd.) I would like to read from a Wikidata item, and then feed it into this template. Is that possible? &mdash; Martin (MSGJ · talk) 22:43, 1 February 2021 (UTC)
 * The following works, but what if P1343 had several values? &mdash; Martin (MSGJ · talk) 22:47, 1 February 2021 (UTC)

→


 * I'm not sure how to do that with Module:Wd, but Module:WikidataIB has some useful functions. Starting from :
 * That just returns the first match, but it can return all matches:
 * So if I arrange for the CiteQ sandbox to accept a list of qids:
 * we can get multiple cites.
 * I can arrange for further parameters to wrap each one in  tags or whatever if required, or perhaps have the cites separated by a selectable separator – I'm just using  for now . See what you think. --RexxS (talk) 02:46, 2 February 2021 (UTC)
 * This could potentially be useful for a Bibliography section, even more so if we can get the multiple values separated by a line break and an asterisk, to produce a bulleted list. ~ ★ nmaia d 03:31, 2 February 2021 (UTC)
 * Agreed! For my use, I would like them inside ref tags, but I can see the use of a bulleted list also. &mdash; Martin (MSGJ · talk) 21:03, 2 February 2021 (UTC)
 * I can arrange for further parameters to wrap each one in  tags or whatever if required, or perhaps have the cites separated by a selectable separator – I'm just using  for now . See what you think. --RexxS (talk) 02:46, 2 February 2021 (UTC)
 * This could potentially be useful for a Bibliography section, even more so if we can get the multiple values separated by a line break and an asterisk, to produce a bulleted list. ~ ★ nmaia d 03:31, 2 February 2021 (UTC)
 * Agreed! For my use, I would like them inside ref tags, but I can see the use of a bulleted list also. &mdash; Martin (MSGJ · talk) 21:03, 2 February 2021 (UTC)

I've implemented a definable separator (defaults to comma space): However, you can't just add a newline and an an asterisk because the parser has already done the list markup before it renders the module output. It's also not good list markup, so I've implemented the ability to pass the multiple cites through any template of your choosing. For example, Template:Unbulleted list (ubl): Another example, Template:Bulleted list: As requested, I've also implemented a tag parameter, which will wrap each cite in the tag supplied. At present, it only recognises ref, and it doesn't implement named refs because I can't see a robust naming convention for now (maybe author1 year1,2,3 … ? but they don't always exist). Anyway here's an example:

Does that give you something to play with? (I mean test thoroughly) --RexxS (talk) 21:31, 3 February 2021 (UTC)
 * I will definitely have a play around with that, thanks. Without a ref name does that mean that the same references may be listed separately multiple times? I wonder how Module:Wd handles the references because it does a good job of grouping them. Anyway I will see what happens. &mdash; Martin (MSGJ · talk) 21:36, 3 February 2021 (UTC)
 * Hmm yes, that is a bit of a problem. On List of lighthouses in China (before I reverted) I had about 20 of the same reference all listed separately. And I notice that when is not present, the template will cite its own Wikidata item which was unexpected, e.g.

→
 * I think I would prefer a blank output if P1343 is not used. &mdash; Martin (MSGJ · talk) 21:47, 3 February 2021 (UTC)
 * Surely the obvious choice of ref name is the Qid of the reference? It doesn't need to be human readable because it will not appear in the wikitext. (But I would suggest to have a look in Module:Wd because it does handle references well ...) &mdash; Martin (MSGJ · talk) 22:09, 3 February 2021 (UTC)
 * I don't see an easy way to group references as I'm still looking for a robust naming scheme.
 * Yes, followQid is designed for use where it has to return a qid into another call, so it returns the input qid if it can't follow the property value to another qid. You should probably use getPropertyIDs, which works more like getValue and returns nothing if there's no :
 * See if that works better for you. --RexxS (talk) 22:16, 3 February 2021 (UTC)
 * Tried it here. Working pretty well! It would be nice if this functionality was further baked into Cite Q so that we'd need to write fewer things. I think if most users have to manually input something as long as, they might as well just cite the books manually. ~ ★ nmaia d  22:20, 3 February 2021 (UTC)
 * Yup, but I was assuming it was going to be used inside a template or infobox, so that it just gets a single qid passed by the editor (or derived from the page). --RexxS (talk) 22:37, 3 February 2021 (UTC)
 * Module:Wd doesn't do anything to group the refs; it just gives each one a name depending on its content. So all the ones with the same content get the same name and the wiki-parser does the grouping. I've implemented your suggestion to use the qid as the name (many thanks for the insight!). Perhaps you could try it out and see if that is better? --RexxS (talk) 23:03, 3 February 2021 (UTC)
 * That is looking pretty good on List of lighthouses in China now! I agree with NMaia that it would be nice if the syntax was a bit nicer, or if Cite Q could be used more generally to display references for statements on Wikidata. But that is a great improvement for now &mdash; Martin (MSGJ · talk) 12:45, 4 February 2021 (UTC)
 * You can get a list of QIDs with Module:Wd as well, simply like this: . You can even customise the separators if you'd like:    Thayts   •••  17:41, 10 February 2021 (UTC)
 * Yup, but I was assuming it was going to be used inside a template or infobox, so that it just gets a single qid passed by the editor (or derived from the page). --RexxS (talk) 22:37, 3 February 2021 (UTC)
 * Module:Wd doesn't do anything to group the refs; it just gives each one a name depending on its content. So all the ones with the same content get the same name and the wiki-parser does the grouping. I've implemented your suggestion to use the qid as the name (many thanks for the insight!). Perhaps you could try it out and see if that is better? --RexxS (talk) 23:03, 3 February 2021 (UTC)
 * That is looking pretty good on List of lighthouses in China now! I agree with NMaia that it would be nice if the syntax was a bit nicer, or if Cite Q could be used more generally to display references for statements on Wikidata. But that is a great improvement for now &mdash; Martin (MSGJ · talk) 12:45, 4 February 2021 (UTC)
 * You can get a list of QIDs with Module:Wd as well, simply like this: . You can even customise the separators if you'd like:    Thayts   •••  17:41, 10 February 2021 (UTC)

This seems like an improper citation. Maybe I missed something, but please explain how this edit is a proper citation. It adds footnote number 36, which claims that the data about the Ponta de Ka-ho Lighthouse is supported by Admiralty List of Lights and Fog Signals. But when I click on the title, I don't go to a web version of Admiralty List of Lights and Fog Signals; I land on a page that describes that publication, and gives me an opportunity to buy it.

So, MSGJ, did you read the entry for Ponta de Ka-ho Lighthouse in Admrialty List of Lights and Fog Signals and verify that it agrees with at least some of the data in the Wikipedia article? If not, then this citation violates the verifiability policy and you should revert it. Jc3s5h (talk) 15:50, 4 February 2021 (UTC)
 * (This is a question to raise at the article's talk page, not here. But anyway ...) There is certainly information about that lighthouse in that publication. The direct link is which is also shown in a different column in that table. I may be able to tweak the footnote so it points to the precise page, but that doesn't invalidate the whole reference. &mdash; Martin (MSGJ · talk) 17:15, 4 February 2021 (UTC)
 * Correction: I see you are referring to the Admiralty list (I misread your diff). No I have not been able to check that one as it is behind a paywall. Its entry is cross-referenced in the NGA list (link supplied above) but I would be happy to remove that reference if you think it is inappropriate. &mdash; Martin (MSGJ · talk) 17:29, 4 February 2021 (UTC)
 * I believe that when an editor adds a reference, the editor is stating the editor has read the reference and the reference supports the claims in the article. The only defensible exception I could imagine is if the editor is translating an article from a different language Wikipedia, and just translates what the article says, without obtaining and verifying each reference. So I believe the Admiralty list citation should be removed.
 * This whole approach seems suspect to me. By adding one is claiming that all sources listed under the  for Ponta de Ka-ho Lighthouse support the claims in the Wikipedia article, even sources that are added in the future. Citations should not be added to articles unless they have been read and the adding editor verifies that the source supports the claims in the article. It should be impossible for a new source to magically appear with no editor looking at it to see if the statements in the source agree with the statements in the Wikipedia article. Jc3s5h (talk) 17:56, 4 February 2021 (UTC)
 * I don't believe that adding content that includes a reference is necessarily a guarantee that the editor has read the source cited. It is quite common for content to be copied from one part of an article to another, or from one article to another, within the allowance of the CC-BY-SA licence. It is normal for such additions to be done on the assumption of good-faith, otherwise the requirement for an editor who copies content to read every source and ensure it supports the nearby content would be too onerous. What is expected by WP:V is that when an editor adds a citation to challenged content, then it must the case that the editor has read the source they are adding. But that is not the case here. Comparing with, you will see that the additions were not made to challenged content.
 * As for the old chestnut that content that is taken from Wikidata can't be used on Wikipedia because somebody might change it on Wikidata, that line of argument has been rejected numerous times already. If someone manually adds paper sources to Wikipedia, we don't have this assumption of bad faith that is being levelled at the addition of the same sources on Wikidata. --RexxS (talk) 18:27, 4 February 2021 (UTC)
 * Something has changed. When I first looked at the lighthouse list article, the "described by source" items from the Wikidata item were shown explicitly in the Wikipedia article. Now they're not.
 * I also now see that this isn't really about Cite Q, it's about list item. So an item is just being copied from Wikidata to a list item in Wikipedia. That approach has been more-or-less accepted. I think the items being copied are disappointing, because the individual statements do not have references; there are just described by source properties, which is roughly equivalent to general references in Wikipedia articles. Jc3s5h (talk) 19:13, 4 February 2021 (UTC)
 * Yes I reverted that change based on your comments. I personally believe that it is a valuable source despite being behind a paywall, but I am erring on the side of caution and taking your concerns seriously. Regarding the sourcing, I am equalling or bettering the levelling of sourcing in the article prior to conversion. In many cases, sources were provided on Wikidata which were not on the article prior, so I can claim a net positive in sourcing. &mdash; Martin (MSGJ · talk) 19:49, 4 February 2021 (UTC)
 * There's lots of sources that I'd like to have, but don't want to spend the money to buy them. Unless someone decides to give them to me as a present, I won't be mentioning them on Wikipedia or Wikidata. Jc3s5h (talk) 19:56, 4 February 2021 (UTC)
 * The same goes for me, but the beauty of working on a massively collaborative project is that somebody, somewhere might have a copy of the source and can verify it. It's a shame that subtle vandalism can weaken trust in that process, but I have to work on the assumption that our editors are all contributing for the benefit of the project. Cheers --RexxS (talk) 22:25, 4 February 2021 (UTC)

Notification if Cite Q is used or if a Cite Q target changes in Wikidata
I would like to set up two varieties of notification. So... is there a way to "patrol" wikidata's Cite Q items? Trilotat (talk) 05:16, 11 February 2021 (UTC)
 * 1) I wouldn't mind knowing when someone uses Cite Q.
 * 2) I'd like to know if the Cite Q reference changes. Maybe we need to add author-links if someone disambiguates an author (and if that author has a Wikipedia article). Maybe we need to repair Wikidata item that disrupts the wikipedia article somehow. I can star the Wikidata items as I Cite Q them, but I would like to keep an eye on the others, too.
 * I don't know any way of being notified when CiteQ is added to an article.
 * It would be possible to add a tracking category to the template, something like Category:Cite Q, although We tend not to create potentially unlimited scale categories unless they have a rationale for their existence. Optionally, you can search for  or see 'What links here" from Template:Cite Q to see articles where it is used.
 * As far as I know, there's no way of knowing whether a Wikidata item changes without checking a Wikidata watchlist. You can enable your English Wikipedia watchlist to show Wikidata changes, but only for articles that are connected to Wikidata (and most Wikidata citations don't have corresponding enwiki articles). --RexxS (talk) 17:30, 11 February 2021 (UTC)
 * Thank you. I searched for hastemplate:"Cite Q" and I get nearly 42.5K results. Checked the first result and it doesn't include Cite Q. I used the advanced search tool, and it doesn't useful info either. Trilotat (talk) 23:08, 11 February 2021 (UTC)
 * if you look at "What links here" for Template:Cite Q, the first article is Azincourt. That article uses Template:Infobox French commune and Template:France metadata Wikidata, which uses Template:Wikidata, which uses Module:Wd, which can call Template:Cite Q to render references. The searches we use will find indirect uses of CiteQ as well as where it is directly used in an article. Hope that makes sense. --RexxS (talk) 15:33, 12 February 2021 (UTC)
 * if you look at "What links here" for Template:Cite Q, the first article is Azincourt. That article uses Template:Infobox French commune and Template:France metadata Wikidata, which uses Template:Wikidata, which uses Module:Wd, which can call Template:Cite Q to render references. The searches we use will find indirect uses of CiteQ as well as where it is directly used in an article. Hope that makes sense. --RexxS (talk) 15:33, 12 February 2021 (UTC)

et al is missing when using author=
I've got a Cite Q reference that has about 10 authors. It automatically truncates to 3 authors with the remaining as et al. If I use author, author2, author 3, et disappears. How can I use Cite Q, specify how the first three authors display, and retain et al for remaining? Trilotat (talk) 15:53, 16 February 2021 (UTC)
 * I resolved it by using authorn for the remaining authors. Trilotat (talk) 16:02, 16 February 2021 (UTC)