Help talk:Citation Style 1

For more talk pages, see Special:WhatLinksHere/Help talk:Citation Style 1.

• Help talk:Citation Style 1 (active, central discussion)
• Help talk:Citation Style 1/accessdate
• Help talk:CS1 errors (active)
• Template talk:Citation (the only Citation style 2 template)
• Template talk:Cite AV media notes
• Template talk:Cite AV media
• Template talk:Cite book
• Template talk:Cite conference
• Template talk:Cite DVD-notes
• Template talk:Cite encyclopedia
• Template talk:Cite episode
• Template talk:Cite interview
• Template talk:Cite journal
• Template talk:Cite mailing list
• Template talk:Cite map
• Template talk:Cite manual
• Template talk:Cite music release notes
• Template talk:Cite news
• Template talk:Cite newsgroup
• Template talk:Cite podcast
• Template talk:Cite press release
• Template talk:Cite report
• Template talk:Cite serial
• Template talk:Cite sign
• Template talk:Cite speech
• Template talk:Cite techreport
• Template talk:Cite thesis
• Template talk:Cite web
• Template talk:Citation Style documentation
• Module talk:CS1
• Module talk:Citation/CS1
• Module talk:Citation/CS1/COinS (active)
• Module talk:Citation/CS1/Configuration
• Module talk:Citation/CS1/Date validation
• Module talk:Citation/CS1/Identifiers
• Module talk:Citation/CS1/styles.css
• Module talk:Citation/CS1/Suggestions
• Module talk:Citation/CS1/Utilities
• Module talk:Citation/CS1/Whitelist
• Module talk:Citation
• Category talk:CS1 errors: dates/Archive 1

The use of accessdate when a page is archived or dead

What is the appropriate use of accessdate when a page is archived or dead? My initial feeling was that the accessdate applies to the archival page (if it exists) and otherwise to the "main" page which has been archived or is dead. Is there guidance somewhere? --Izno (talk) 00:34, 19 February 2013 (UTC)

Access date is simply the date the link was accessed, usually the date it was added. There is no requirement to include it, and it can be hidden by registered users. It is mainly useful for a web page that changes frequently, or when there is no discernible publication date. --— Gadget850 (Ed) ^talk 02:06, 19 February 2013 (UTC)

If the page is dead, and there is an access date, the latter apparently can serve to locate an archive version that most closely matches the date the citation was placed, but such utility is IMHO debatable. If there is an online archive of the article, I think it makes little sense to leave an access date that serves a purpose no longer. -- Ohconfucius ^{ping / poke} 02:22, 19 February 2013 (UTC)

I believe I'm the one who argued that, in fact, and applied just that logic today in cleaning up some citations. :) I might agree with the latter sentence, but I'd like thoughts on the below question before I do so. --Izno (talk) 04:54, 19 February 2013 (UTC)

Yes, but for which link? The archive link or the archived link? I suppose it seems a little incredulous to expect the archive link to go down... --Izno (talk) 04:53, 19 February 2013 (UTC)

however, it does happen (at least temporarily). i always archive online sources (when allowed), preferably at WebCite, and have encountered service disruptions several times. 70.19.122.39 (talk) 14:06, 19 February 2013 (UTC)

In my opinion: when the original link is dead, then there is no need for an access date, as the archive has an archive date that will normally not change. Where the original link is still live then the access date is useful if there is no publication date, regardless of whether it is preemptively archive. --— Gadget850 (Ed) ^talk 10:57, 19 February 2013 (UTC)

If you are adding an |accessdate=, it pertains to the original |url=. The |archivedate= (although being the date that the archiving service grabbed the page and not the date when you viewed it) does a similar job for |archiveurl= - we know that the archived page isn't going to change further. --Redrose64 (talk) 13:30, 19 February 2013 (UTC)

Yes, but our ability to access the archived page might [change]. That's the only Pandora's box I'm opening here. :) If the lesser consensus here disagrees with me, that's fine, and I'll remove (or at least, not add) accessdates when the cited pages are archived. To that end, I'm not sure the below proposal captures the point I'm making. --Izno (talk) 18:09, 19 February 2013 (UTC)

Documentation

This has come up before, so we should tweak the documentation. Current:

accessdate: Full date when URL was accessed; use the same format as other access and archive dates in the citations; do not wikilink. Can be hidden by registered editors.

Proposed:

accessdate: Full date when original URL was accessed; use the same format as other access and archive dates in the citations; do not wikilink. Not required for web pages that do not change; mainly of use for web pages that change frequently or have no publication date. Can be hidden by registered editors.

--— Gadget850 (Ed) ^talk 15:00, 19 February 2013 (UTC)

the reflink doesn't link to a ref here, do you mean the note at the template page? other than that i agree. 70.19.122.39 (talk) 15:13, 20 February 2013 (UTC)

I meant to leave that out for this. It links to an explanatory note on date formats. --— Gadget850 (Ed) ^talk 15:18, 20 February 2013 (UTC)

Could we maybe add a bit stronger guidance, something along the lines of "use of accessdate is [weakly] discouraged where the page has an archive and associated archive date, but is allowed where desired", per the above discussion? --Izno (talk) 22:11, 27 February 2013 (UTC)

 Done --— Gadget850 (Ed) ^talk 12:44, 14 March 2013 (UTC)

ORCID

ORCID identifiers disambiguate journal contributors with similar names; and unite records for authors who write under more than one name. Think of them as an ISSN for people. For example. my ORCID is 0000-0001-5882-6823 and the corresponding URI is http://orcid.org/0000-0001-5882-6823

I believe that we should add an option for an ORCID parameter for each author of a cited work, to CS1 templates. Whether or not to make that a clickable link can be debated. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:33, 8 March 2013 (UTC)

I believe ORCIDs are a good idea within the {{Authority control}} template on Wikipedia articles about individual researchers. I am less convinced that we should clutter our bibliographies with them, especially in cases where we have a wikilink to the author's article. —David Eppstein (talk) 22:17, 8 March 2013 (UTC)

I suspect (and would welcome evidence) that in the vast majority of cases, we have no article about the author, and thus no such link. ORCIDs would not be clutter, but valuable information. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:53, 9 March 2013 (UTC)

I support this. It should be a clickable link and the template should check the format of the ORCID for mangled ones. Stuartyeates (talk) 06:12, 9 March 2013 (UTC)

I agree with David, they should be in the authority control system and not in the citation templates themselves. I don't see how they help a reader locate a copy of the source to verify the information being cited. I do see how the ISSN would help a reader locate a library that maintains a copy of the journal/magazine, but I'm not sure the same can be said for ORCIDs. Imzadi 1979 → 06:37, 9 March 2013 (UTC)

You're quite right. But then no such claim has been made. An ORCID doesn't help find the publication; it helps find the - and unambiguously to identify - the author; and thus to find that author's other works, affiliation, reputation, etc. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:51, 9 March 2013 (UTC)

The purpose of a citation is to allow a reader to find the publication being cited for verification purposes. The other works of the author, etc would best be spelled out in an article on that author, which conveniently enough, |authorlink= would provide. From that article then, the ORCID and other authority control information would be linked. Anything else in the citation is clutter. That isn't to say that the ORCID isn't valuable, just that it isn't needed, in this context. Imzadi 1979 → 12:31, 9 March 2013 (UTC)

Please refer to my above comment, where I have already addressed the |authorlink= red herring. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 12:35, 9 March 2013 (UTC)

one solution would be to have "authorlink" output either a wikilink or an ORCID? add logic in the template so that if the field's data start with a number link to ORCID. i realize there could be situations where people (legally) use numbers as names. in which case my idea would not be optimal. 70.19.122.39 (talk) 15:37, 9 March 2013 (UTC)

If we introduce data like this (sounds useful for cross referencing, metadata), we have to be able to strike a balance between the additional data and a default display of the reference that is not too detailed for an average reader. Not sure how we would do this/how it would be configured though. Rjwilmsi 10:17, 12 March 2013 (UTC)

If we must, we could hide ORCID numbers behind an icon, or the text "ORCID"; but displaying them would be better. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:31, 12 March 2013 (UTC)

.39's solution would be technically easy, just "if authrolink exists, do not display ORCID". Numeric names would not be a problem. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:31, 12 March 2013 (UTC)

your amendment is more elegant and takes care of names that are numbers. the only potential problem i can see is that #ifexist is expensive, or more so than plain #if statements that don't check article namespace. conceivably an article could bump into limits when there are many citation templates? 70.19.122.39 (talk) 14:00, 12 March 2013 (UTC)

And it would mean that when printed or ported, the ORCID would be hidden. --— Gadget850 (Ed) ^talk 14:15, 12 March 2013 (UTC)

That's easily dealt with by CSS, as we do with external links in citations, for instance. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:22, 12 March 2013 (UTC)

Doesn't Lua resolve this? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:22, 12 March 2013 (UTC)

Would someone like to knock up a test implementation in a sandbox, please? If you need a sample ORCID, mine is on my user page. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:26, 12 March 2013 (UTC)

Anyone? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:19, 22 March 2013 (UTC)

I support this conceptually but am not sure how to participate or what is next to be done. Blue Rasberry (talk) 14:35, 27 March 2013 (UTC)

RESOLVED: Use current "postscript=" for any additional data: The current wp:CS1 format allows parameter "postscript=. ORCID [http://orcid.org/0000-0001-9977-9999 9977-9999]." to add the ORCID id, or any of millions of other extra data items into each citation. That option has been available for years, so feel free to discuss with other editors of each article when wanting to append more details. The postscript parameter can accept over 50,000 characters of data, including maps with driving instructions to the library, to help readers find each source document. -Wikid77 (talk) 20:49, 27 March 2013 (UTC)

NOT RESOLVED and please don't unilaterally declare your suggestion as a resolution. Using |postscript= would mean that an ORCID identity would not be identified as such; not automatically linked to the relevant URL; and that its display would not be suppressed in the presence of an |authorlink=, as discussed above. Your map comment is asinine. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:04, 27 March 2013 (UTC)

Cool your jets. Wikid77 wasn't saying that a resolution had been reached, but rather was following the quaint forensic tradition of prefacing a proposed decision with RESOLVED i.e. it means I move that it be RESOLVED that.... See [1]. It's a peculiar idiom indeed, but one that, to be frank, should be well within the experience of anyone dealing with the niceties of research and citation, so please have a care in future before flying off the handle. An apology seems in order. EEng (talk) 23:17, 27 March 2013 (UTC)

Um, no. That may well be what was in Wikid77's head, but this page uses the English language, not some "peculiar", "quaint forensic tradition". You don't see that usage elsewhere on Wikipedia, nor even on the rest of this page. It's not unreasonable to suppose that other readers will see the word "resolved", emboldened and in all-caps, and pass on to the next section (c.f. {{resolved}}). Now, can anyone help with the ORCID parameter requested? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:01, 28 March 2013 (UTC)

There is much that is peculiar or quaint which may reasonably be expected to be within the experience of the culturally literate. Apparently you've never served on a board of directors, public commission, or student council; participated in any kind of formal discussion or debate; been present at a faculty or school board meeting, or a union vote; owned stock; or read about any of these things with any depth of comprehension. That's OK, of course; what's not OK is that, faced with an opportunity to expand your knowledge just a bit, you choose instead to insist that the rest of us restrict our discourse to your cramped radius of experience. It's kind of like the old joke about opposition to foreign-language instruction in the schools: "If English was good enough for Jesus, it's good enough for our kids!" EEng (talk) 21:48, 28 March 2013 (UTC)

Well, that's quite a collection of attacks and failures to AGF. No wonder you leap to so many false conclusions. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:14, 29 March 2013 (UTC)

• 'postscript' is for the terminating punctuation of the citation- please don't suggest misusing it. I am sure some editors are doing something odd and wonderful, but we have already broken a few misused parameters. An it won't work for ORCID as there will be one ORCID per author and it has to be attached to the author.
• Putting the ORCID on the author page and linking to it with 'authorlink' will not work if the page is ported to another wiki but the author page is not.
• The {{authority control}} template is for the subject of the article, not for authors of individual sources used as citations. ORCID would be good there for use when the article is about a specific author
• The key question is: does ORCID help identify the source? As I see it, ORCID does help to uniquely identify the author, regardless of duplicate names or name changes.
• But, ORCID is not the only author identifier. It is actually a subset of the International Standard Name Identifier which further allows for pseudonyms and publisher imprints. ORCID is for researchers and academic authors, whereas ISNI covers music, film and other media. I think we should use both. I would also like to see if they can both be included in one parameter.
• As noted, ORCD/ISNI would be attached to individual author names.
• We are in the process of updating CS1 templates to Lua. Lets add this to the queue as a new feature while I look at adding this to {{citation/core}}. --— Gadget850 (Ed) ^talk 13:55, 28 March 2013 (UTC)

We recently discussed the issue of whether to store ORCID and ISNI as one or two parameters during a Wikidata IRC session and tended to the latter; transcript with reasoning at d:Wikidata:Requests for permissions/Bot/VIAFbot/Meeting agenda. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:37, 1 April 2013 (UTC)

This should go to Module talk:Citation/CS1/Feature requests. Please clarify ORCID v. ISNI. --  Gadget850 (Ed) ^talk 11:25, 7 April 2013 (UTC)

|publication-date= does not pass data to {{harv}} when verbose

pertinent examples are emphasized. i suspect whitespace is the prob.

Markup	Renders as
{{harvnb|Author|2000}} {{harvnb|Author|2001}} {{strong|{{harvnb|Author|2002}}}} {{strong|{{harvnb|Author|2003}}}} {{harvnb|Author|2004}} (Bibliography) {{refbegin}} * {{cite book|ref=harv|last=Author|date=1 January 2000|title=Title}} * {{cite book|ref=harv|last=Author|title=Title|publication-date=2001}} * {{strong|{{cite book|ref=harv|last=Author|title=Title|publication-date=January 2002}}}} * {{strong|{{cite book|ref=harv|last=Author|title=Title|publication-date=1 January 2003}}}} * {{cite book|ref={{harvid|Author|2004}}|last=Author|title=Title|publication-date=1 January 2004}} {{refend}}	Author 2000 Author 2001 Author 2002 Author 2003 Author 2004 (Bibliography) Author (1 January 2000). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help) Author (2001). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help) Author (January 2002). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help) Author (1 January 2003). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help) Author (1 January 2004). Title. {{cite book}}: |last= has generic name (help)CS1 maint: ref duplicates default (link)

70.19.122.39 (talk) 14:24, 13 March 2013 (UTC)

The five {{harvnb}} are generating links to, in order, href="#CITEREFAuthor2000" href="#CITEREFAuthor2001" href="#CITEREFAuthor2002" href="#CITEREFAuthor2003" href="#CITEREFAuthor2004" whereas the generated anchors are id="CITEREFAuthor2000" id="CITEREFAuthor2001" id="CITEREFAuthorJanuary_2002" id="CITEREFAuthor1_January_2003" id="CITEREFAuthor2004" But the |harv=ref code only uses |publication-date= as a fallback, when both |year= and |date= are absent:

Markup	Renders as
{{harvnb|Author|2010}} {{harvnb|Author|2011}} {{strong|{{harvnb|Author|2012}}}} {{strong|{{harvnb|Author|2013}}}} {{harvnb|Author|2014}} (Bibliography) {{refbegin}} * {{cite book|ref=harv|last=Author|date=1 January 2010|title=Title}} * {{cite book|ref=harv|last=Author|title=Title|date=2011-01-01}} * {{strong|{{cite book|ref=harv|last=Author|title=Title|date=January 2012}}}} * {{strong|{{cite book|ref=harv|last=Author|title=Title|date=1 January 2013}}}} * {{cite book|ref={{harvid|Author|2014}}|last=Author|title=Title|date=1 January 2014}} {{refend}}	Author 2010 Author 2011 Author 2012 Author 2013 Author 2014 (Bibliography) Author (1 January 2010). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help) Author (2011). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help) Author (January 2012). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help) Author (1 January 2013). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help) Author (1 January 2014). Title. {{cite book}}: |last= has generic name (help)CS1 maint: ref duplicates default (link)

|publication-date= is intended for use in addition to, not instead of, |date=, so when you only have one, it's best to use |date=. --Redrose64 (talk) 15:30, 13 March 2013 (UTC)

Correct: It is the date of publication when different from the date the work was written. A real example:

White, T.H. (1941). The Book of Merlyn. University of Texas Press (published 1977). {{cite book}}: Invalid |ref=harv (help)

--— Gadget850 (Ed) ^talk 16:00, 13 March 2013 (UTC)

i think you guys misunderstood.

|publication-date= passes data to hrefs when (a) only the year is given (b) the href is input with date formated exactly as the data in the "publication-date" field, or (c) {{harvid}} is used

Markup	Renders as
(a){{harvnb|Author3|2012}} * {{cite book|ref=harv|last=Author3|publication-date=2012|title=Title}} (b){{harvnb|Author4|Jan 1, 2012}} * {{cite book|ref=harv|last=Author4|publication-date=Jan 1, 2012|title=Title}} (c){{harvnb|Author5|2012}} * {{cite book|ref={{harvid|Author5|2012}}|last=Author5|publication-date=Jan 1, 2012|title=Title}}	(a) Author3 2012 Author3 (2012). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help)CS1 maint: numeric names: authors list (link) (b) Author4 & Jan 1, 2012 Author4 (Jan 1, 2012). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help)CS1 maint: numeric names: authors list (link) (c) Author5 2012 Author5 (Jan 1, 2012). Title. {{cite book}}: |last= has generic name (help)CS1 maint: numeric names: authors list (link) CS1 maint: ref duplicates default (link)

when a citation is formatted with a verbose date (eg Jan 1, 2012) template code tries to extract year from |date=, and this is passed to |ref=harv.

Markup	Renders as
{{harvnb|Author1|2012}} *{{cite book|ref=harv|last=Author1|date=Jan 1, 2012|title=Title}}	Author1 2012 Author1 (Jan 1, 2012). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help)CS1 maint: numeric names: authors list (link)

but this does not happen with |publication-date=.

Markup	Renders as
{{harvnb|Author2|2012}} *{{cite book|ref=harv|last=Author2|publication-date=Jan 1, 2012|title=Title}}	Author2 2012 Author2 (Jan 1, 2012). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help)CS1 maint: numeric names: authors list (link)

this is a bug. pls fix.

the suggestion to use |date= when only the publication date is known presents several problems.

what happens in online sources (eg {{cite web}}) when neither work date nor pub. date are known? do we input the access date in |date=?

if not, there's an inconsistency. [when no other date is known, and for verification purposes (the reason citations exist) the access date is the defacto "publication date", which should then subst into |date= per your argument].

on the other hand, if the access date should become the variable data for |date= then the dependency between |accessdate= and |url= should be undone to avoid errors.

and what about archival dates when |deadurl=yes? another clarification should be provided.

secondly, |publication-date= has obvious semantic significance for editors/citation providers. substitution with |date= should not be encouraged. instead |publication-date= should be promoted to the same status as |date= within the template when |date= is absent. similarly for online sources |accessdate= should be promoted when |date= and |publication-date= is absent. and so on for |archivedate=.

or, just do the lazy thing and remove |publication-date= altogether from CS1.

70.19.122.39 (talk) 00:19, 14 March 2013 (UTC)

Yes, it does not happen with |publication-date= but it is not a bug; it is a side-effect of the non-use of other parameters. Quite frankly I don't know why we attempt to extract a year from |publication-date= in the first place, since |publication-date= (see documentation) is intended to be supplementary to either |date= or |year= (see documentation).

Access dates must only be placed in the |accessdate= parameter (see documentation), no other parameter is intended for this.

The |ref= parameter (see documentation) is pretty much free-form. It is provided so that a manually-constructed anchor may be attached to the {{cite book}} when the special value |ref=harv does not give a suitable anchor. Use of the {{harvid}} template is not obligatory: it is a tool for constructing an anchor which is consistent with the links generated by templates like {{harv}}, {{harvnb}} and {{sfn}}. --Redrose64 (talk) 10:58, 14 March 2013 (UTC)

I need to fix the doc for this parameter. Somehow I thought the field did not display if year or date were not defined. --— Gadget850 (Ed) ^talk 12:46, 14 March 2013 (UTC)

No, that is how it works. If publication-date but not date, then publication-date is used as date and the anchor is formed:

Markup	Renders as
{{cite book |last=White |first=T.H. |authorlink=T. H. White |title=The Book of Merlyn |date=1941 |publication-date=1977 |publisher= University of Texas Press |ref=harv}}	White, T.H. (1941). The Book of Merlyn. University of Texas Press (published 1977). {{cite book}}: Invalid |ref=harv (help) '"`UNIQ--templatestyles-00000079-QINU`"'<cite id="CITEREFWhite1941" class="citation book cs1">[[T. H. White|White, T.H.]] (1941). ''The Book of Merlyn''. University of Texas Press (published 1977).</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=The+Book+of+Merlyn&rft.pub=University+of+Texas+Press&rft.date=1941&rft.aulast=White&rft.aufirst=T.H.&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1" class="Z3988"></span> <span class="cs1-visible-error citation-comment"><code class="cs1-code">{{[[Template:cite book|cite book]]}}</code>: </span><span class="cs1-visible-error citation-comment">Invalid <code class="cs1-code">&#124;ref=harv</code> ([[Help:CS1 errors#invalid_param_val|help]])</span>
{{cite book |last=White |first=T.H. |authorlink=T. H. White |title=The Book of Merlyn |publication-date=1977 |publisher= University of Texas Press |ref=harv}}	White, T.H. (1977). The Book of Merlyn. University of Texas Press. {{cite book}}: Invalid |ref=harv (help) '"`UNIQ--templatestyles-0000007D-QINU`"'<cite id="CITEREFWhite1977" class="citation book cs1">[[T. H. White|White, T.H.]] (1977). ''The Book of Merlyn''. University of Texas Press.</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=The+Book+of+Merlyn&rft.pub=University+of+Texas+Press&rft.date=1977&rft.aulast=White&rft.aufirst=T.H.&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1" class="Z3988"></span> <span class="cs1-visible-error citation-comment"><code class="cs1-code">{{[[Template:cite book|cite book]]}}</code>: </span><span class="cs1-visible-error citation-comment">Invalid <code class="cs1-code">&#124;ref=harv</code> ([[Help:CS1 errors#invalid_param_val|help]])</span>
{{cite book |last=White |first=T.H. |authorlink=T. H. White |title=The Book of Merlyn |publication-date=1 May 1977 |publisher= University of Texas Press |ref=harv}}	White, T.H. (1 May 1977). The Book of Merlyn. University of Texas Press. {{cite book}}: Invalid |ref=harv (help) '"`UNIQ--templatestyles-00000081-QINU`"'<cite id="CITEREFWhite1977" class="citation book cs1">[[T. H. White|White, T.H.]] (1 May 1977). ''The Book of Merlyn''. University of Texas Press.</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=The+Book+of+Merlyn&rft.pub=University+of+Texas+Press&rft.date=1977-05-01&rft.aulast=White&rft.aufirst=T.H.&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1" class="Z3988"></span> <span class="cs1-visible-error citation-comment"><code class="cs1-code">{{[[Template:cite book|cite book]]}}</code>: </span><span class="cs1-visible-error citation-comment">Invalid <code class="cs1-code">&#124;ref=harv</code> ([[Help:CS1 errors#invalid_param_val|help]])</span>

This should work. Let me look at the original examples again. --— Gadget850 (Ed) ^talk 13:23, 14 March 2013 (UTC)

Here is what is happening: {{harvnb}} is including the year only in the link and {{cite book}} is including the full date in the anchor, causing a mismatch:

Markup	Renders as
{{harvnb|Author|2013}} {{cite book|ref=harv|last=Author|publication-date=Jan 1, 2013|title=Title}}	Author 2013 Author (Jan 1, 2013). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help) [[#CITEREFAuthor2013|Author 2013]]<span class="error harv-error" style="display: inline; font-size:100%"> harvnb error: multiple targets (5×): CITEREFAuthor2013 ([[:Category:Harv and Sfn template errors|help]])</span> '"`UNIQ--templatestyles-00000088-QINU`"'<cite id="CITEREFAuthor2013" class="citation book cs1">Author (Jan 1, 2013). ''Title''.</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=Title&rft.date=2013-01-01&rft.au=Author&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1" class="Z3988"></span> <span class="cs1-visible-error citation-comment"><code class="cs1-code">{{[[Template:cite book|cite book]]}}</code>: </span><span class="cs1-visible-error citation-comment"><code class="cs1-code">&#124;last=</code> has generic name ([[Help:CS1 errors#generic_name|help]])</span>; <span class="cs1-visible-error citation-comment">Invalid <code class="cs1-code">&#124;ref=harv</code> ([[Help:CS1 errors#invalid_param_val|help]])</span>

If you include the full date in {{harvnb}}, then it works:

Markup	Renders as
{{harvnb|Author|Jan 1, 2014}} {{cite book|ref=harv|last=Author|publication-date=Jan 1, 2014|title=Title}}	Author & Jan 1, 2014 Author (Jan 1, 2014). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help) [[#CITEREFAuthorJan_1,_2014|Author &amp; Jan 1, 2014]]<span class="error harv-error" style="display: none; font-size:100%"> harvnb error: no target: CITEREFAuthorJan_1,_2014 ([[:Category:Harv and Sfn template errors|help]])</span> '"`UNIQ--templatestyles-0000008D-QINU`"'<cite id="CITEREFAuthor2014" class="citation book cs1">Author (Jan 1, 2014). ''Title''.</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=Title&rft.date=2014-01-01&rft.au=Author&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1" class="Z3988"></span> <span class="cs1-visible-error citation-comment"><code class="cs1-code">{{[[Template:cite book|cite book]]}}</code>: </span><span class="cs1-visible-error citation-comment"><code class="cs1-code">&#124;last=</code> has generic name ([[Help:CS1 errors#generic_name|help]])</span>; <span class="cs1-visible-error citation-comment">Invalid <code class="cs1-code">&#124;ref=harv</code> ([[Help:CS1 errors#invalid_param_val|help]])</span>

At this point, we can debate whether to include the full date in {{harvnb}} or to change the cite templates to include only the year in the anchor. --— Gadget850 (Ed) ^talk 13:43, 14 March 2013 (UTC)

It would be a massive change to put a full date into {{harvnb}} and related templates; people are very much used to using the year alone. As things stand, the cite templates already include only the year in the anchor if you use |date=. It is only |publication-date= where there is inconsistency. See my post of 15:30, 13 March 2013

If we should build the harv anchor from |publication-date=, we should do so in the same manner as |date=. That is, instead of:

|Year={{{year|{{    <!-- attempt to derive year from date, if possible -->
           #if: {{{date|}}}
           |{{
              #iferror:{{#time:Y|{{{date|}}} }}
              |{{#iferror:{{#time:Y|{{{publication-date|einval}}} }}||{{#time:Y|{{{publication-date|}}} }}}}
              |{{#time:Y|{{{date|}}} }}
            }}
           |{{{publication-date|}}} <!-- last resort -->
         }}
      }}}

we should put

|Year={{{year|{{    <!-- attempt to derive year from date, if possible -->
           #if: {{{date|}}}
           |{{
              #iferror:{{#time:Y|{{{date|}}} }}
              |{{#iferror:{{#time:Y|{{{publication-date|einval}}} }}||{{#time:Y|{{{publication-date|}}} }}}}
              |{{#time:Y|{{{date|}}} }}
            }}
           |{{#iferror:{{#time:Y|{{{publication-date|}}} }}||{{#time:Y|{{{publication-date|}}} }}}} <!-- last resort -->
         }}
      }}}

--Redrose64 (talk) 14:05, 14 March 2013 (UTC)

I agree. I just got publication-date added to the Lua version, and it needs some more fixes including the anchor, so this is a good time for that. --— Gadget850 (Ed) ^talk 14:27, 14 March 2013 (UTC)

yes, this is a good solution imo, and goes some way into normalizing date output across CS1. i still have problems with the rationale imposing limits on the use of access date when sources are online (eventually ALL sources may be online). the doc reflects these limitations. imo both should be changed, but that's another issue. Redrose, i don't mean to nitpick, but when software does not perform as expected, eg forming anchors in unpredictable fashion, then it is buggy software. but that's also another matter. 70.19.122.39 (talk) 14:45, 14 March 2013 (UTC)

Now in {{cite book/sandbox}}:

Markup	Renders as
{{harvnb|Author|2015}} {{cite book/sandbox |ref=harv|last=Author|publication-date=Jan 1, 2015|title=Title}}	Author 2015 Author (Jan 1, 2015). Title. {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help) [[#CITEREFAuthor2015|Author 2015]]<span class="error harv-error" style="display: inline; font-size:100%"> harvnb error: multiple targets (2×): CITEREFAuthor2015 ([[:Category:Harv and Sfn template errors|help]])</span> '"`UNIQ--templatestyles-00000095-QINU`"'<cite id="CITEREFAuthor2015" class="citation book cs1">Author (Jan 1, 2015). ''Title''.</cite><span title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=Title&rft.date=2015-01-01&rft.au=Author&rfr_id=info%3Asid%2Fen.wikipedia.org%3AHelp+talk%3ACitation+Style+1" class="Z3988"></span> <span class="cs1-visible-error citation-comment"><code class="cs1-code">{{[[Template:cite book|cite book]]}}</code>: </span><span class="cs1-visible-error citation-comment"><code class="cs1-code">&#124;last=</code> has generic name ([[Help:CS1 errors#generic_name|help]])</span>; <span class="cs1-visible-error citation-comment">Invalid <code class="cs1-code">&#124;ref=harv</code> ([[Help:CS1 errors#invalid_param_val|help]])</span>

This would break any instances where someone used a full date in a harv template, but I would expect this to be nonexistent or rare. Do need to update the anchor documentation to include publication-date. --— Gadget850 (Ed) ^talk 15:03, 14 March 2013 (UTC)

It would only break the harv linking if:

• the harv template had used a full date (the docs explicitly state that the year of publication should be given)
• and in {{cite book}} they had specified
  • |publication-date= with a full date
  • and |ref=harv
  • and had omitted |date= or left it blank
  • and had entirely omitted |year= (not just left it blank)

Pretty big set of ifs. --Redrose64 (talk) 15:50, 14 March 2013 (UTC)

I searched the archives and don't see this was every reported before, so I think it is pretty safe to go forward. --— Gadget850 (Ed) ^talk 16:21, 14 March 2013 (UTC)

also in {{harvs}}:

|Year={{{year1|{{{year|{{    <!-- attempt to derive year from date, if possible -->
             #if: {{{date|}}}
             |{{
                #switch: {{#time:Y|{{{date|}}}}}
                |Error: invalid time = {{{publication-date|}}}
                |{{#time:Y|{{{date|}}}}}
              }}
             |{{{publication-date|}}} <!-- last resort -->
           }}
        }}}}}}

this should be fixed too. you'd think this routine would be in {{citation/core}} since apart from {{harvard citations/core}} and {{cite book}} it applies to {{cite web}}, and any template that uses |publication date=. nah, that'd have made sense. 70.19.122.39 (talk) 00:49, 22 March 2013 (UTC)

Multiple ISBNs

[The key point of this request has been bolded big throughout to -- yes -- shout in desperation how often it's been repeated to no avail.] --EEng (talk) 14:57, 20 March 2013 (UTC)

Is there some way in cite book etc. to indicate multiple or alternative ISBNs e.g. paperback vs hardcover? Thanks. EEng (talk) 21:05, 16 March 2013 (UTC)

This is asked quite often; full answers are in the archives, but in brief: give the ISBN for the single edition/version which you actually consulted. --Redrose64 (talk) 21:37, 16 March 2013 (UTC)

That seems like a cop-out. Part of the purpose of ISBN is to allow the reader to find a copy he can consult for himself, so assuming that e.g. paperback and hardcover have identical content, it's useful to include both. Maybe there's a "misc info" parameter I can just stick the other ISBN into as free text? EEng (talk) 22:59, 16 March 2013 (UTC)

As a source we only need the one version - That said let me help - I do it like the following for bibliographies - not sure its ok but been doing it for 7 years - you dont need ISBN's in the {{cite book}} template itself to work - <ref>ISBN 978-0-8020-5016-8</ref> = ISBN 978-0-8020-5016-8

So this coding that has the ISBN's outside the {{cite book}} but with in the <ref> </ref> parameter will work as a reference.

• <ref>{{cite book |last = Taylor |first = Martin Brook|coauthor= Owram, Doug|year =1994|title =Canadian History|volume=|publisher= University of Toronto Press}} Hard cover pp 12, ISBN 978-0-8020-5016-8 and Paperback pp 24, ISBN 978-0-8020-2801-3</ref>
  

Will render this...

• Taylor, Martin Brook (1994). Canadian History. University of Toronto Press. {{cite book}}: Unknown parameter |coauthor= ignored (|author= suggested) (help) Hard cover pp 12, ISBN 978-0-8020-5016-8 and Paperback pp 24, ISBN 978-0-8020-2801-3

--Moxy (talk) 23:22, 16 March 2013 (UTC)

That's what I was doing -- was hoping for something more elegant. But thanks. EEng (talk) 23:30, 16 March 2013 (UTC)

Except that "pp" is the abbreviation for "pages", and since only one page number is given, it should be Hard cover p. 12, ISBN 978-0-8020-5016-8 and Paperback p. 24, ISBN 978-0-8020-2801-3 --Redrose64 (talk) 23:50, 16 March 2013 (UTC)

And that stuff outside the template will not render metadata in the new Lua templates as we migrate. --— Gadget850 (Ed) ^talk 05:06, 17 March 2013 (UTC)

Millions and millions of entries not in templates - any plans to address this?Moxy (talk) 05:30, 17 March 2013 (UTC)

No. ISBN is not required by any major citation style, and is not completely reliable as a unique identifier. We could provide the capability to add 16 different ISBNs, but it would not directly identify the particular edition used in creating the citation. --— Gadget850 (Ed) ^talk 13:43, 17 March 2013 (UTC)

Pardon my ignorance, but:

• In what way is ISBN not completely reliable as a unique identifier? If you're going to say that there are occasional glitches such as the same number issued twice, or multivolume works assigned a single number, then by that token we might as well not bother with any kind of identifier of anything.
• What does it matter that no major citation style requires ISBN? Here at WP it's recognized as a highly useful thing to include, so it's worth giving attention to making its inclusion as easy as possible.

EEng (talk) 15:07, 17 March 2013 (UTC)

Well, an editor should be specifying the exact edition of a work being referenced. The pagination and typesetting could differ between editions printed in the UK and and the US, for instance. The pagination normally differs between hardcover and paperback editions of the same edition of a book. An editor could be consulting a revised or updated edition as well. Each of these editions will have differing ISBNs, and the principle is to cite the actual source consulted. Imzadi 1979 → 15:58, 17 March 2013 (UTC)

I understand all that, and certainly any provision for multiple ISBNs should include some way to single out the particular ISBN that was consulted, because the core imperative is that the reader be certain he's looking at precisely the work to which the citation refers. It's still useful to include additional ISBNs, with the caveat that different ISBNs may be subject to the sorts of variation you mention. On the other hand, it's often the case the a publisher will explicitly state that two ISBNs are the same page images in different bindings, and it would be useful to be also able to designate those when that's known. EEng (talk) 17:19, 17 March 2013 (UTC)

Good— you have identified the issues with ISBN. JSTOR, doi and others use well verified systems with a central point of authority.

The purpose of a citation is to help the reader identify the source used to make a point in the content. If you have a pressing need to identify versions other than the one used, then put it in a further reading section. Alternate sources that were not consulted do not belong in references. --— Gadget850 (Ed) ^talk 22:38, 17 March 2013 (UTC)

Oh, honestly! Pressing need? It's amazing the amount of verbiage being wasted to characterize as illconceived something that's obviously useful. We've even had one person interject a correction about pp vs p, as if somehow anyone here needed that pointed out.

Let me state the situation again: Frequently a work is issued in paperback and hardcover, utterly identical except for the binding. Sometimes I'll have both on my bookshelf, so it doesn't make sense to talk about "the one actually consulted." It would be stupid to have two separate entries, one for the paperback, one for the hardcover. I could arbitrarily pick one or the other to use in the cite, but that unnecessarily reduces the ease with which the reader can find a copy for his own consultation.

So if you can find it in your heart, please make some way accommodate the entirely sensible desire to have two or more ISBNs in the same entry.

EEng (talk) 05:10, 18 March 2013 (UTC)

The point is that the reader should not be mislead into thinking both books have been verified to support the assertion against which the ref is called out. Hence we would need citations to the intent of:

Author. Title Location:Publisher (1999) ISBN 978-0123456789 pp.3, 5, 7-9. Also available as ISBN 978-9876543210 (not consulted).

Of course the problem then would be how many versions, editions, and printings to enumerate. Instead, we keep it simple and cite just the one consulted. There are many tools, magically linked via wp:Booksources, for finding the others. LeadSongDog come howl! 17:08, 19 March 2013 (UTC)

I don't see the point of insisting on one ISBN "not consulted" and the other as primary when the person adding the citation has checked that both have identical text. Your comment comes across as not having actually read the previous one. Anyway, can't this all be handled by using |id= instead of |isbn=? —David Eppstein (talk) 17:27, 19 March 2013 (UTC)

Well, of course I read it here on this talk page, but how many article readers, or even editors, are likely to do so? If we start down the path of including multiple ISBNs, we will very shortly have many editors showing several ISBNs on the untested premise that they are equivalent. Even in the rare cases where an editor does have the multiple editions at hand and takes the time to check that they are equivalent, it will still make it more difficult for other editors to locate and verify them all. Still, in those rare cases, it would still be possible to say both were consulted, using free-form citations. The template, however, should reflect the usual case as its default. Unless our standard for wp:V is now "one or more of this cited list of editions supports the in-text statement", it would be a move backwards to list multiple editions (even without considering the visual clutter). So yes, EEng, I hold that it is better to just "arbitrarily pick one or the other" than list both. LeadSongDog come howl! 21:27, 19 March 2013 (UTC)

If the only difference between two printings of a book is hard or soft cover, and a statement (etc.) can be verified in both (and possibly the editor has verified in both), then it seems a disservice not to provide as much information as possible. But it seems sufficient that in such cases the additional isbns be put into an {{isbn}} template following the citation, as illustrated above. I am doubtful on having supplemental isbn parameters in the citation templates themselves lest some editors think they need be filled in. And similarly for using |id= for that purpose. If there are two identical (except for the cover) editions, equally consulted, then it shouldn't matter which one is the "master" isbn. ~ J. Johnson (JJ) (talk) 23:08, 19 March 2013 (UTC)

{{ISBN}} doesn't do what you think it does. ISBNs can use magic linking by simply typing ISBN a space and the ISBN. --— Gadget850 (Ed) ^talk 00:36, 20 March 2013 (UTC)

Huh? EEng (talk) 01:23, 20 March 2013 (UTC)

Help:Magic links --— Gadget850 (Ed) ^talk 02:00, 20 March 2013 (UTC)

LSD, I don't know which alternative state of the universe is the sadder one -- David's, in which you didn't bother to parse the use case offered; or your own, incorporating as it does your strained (or, perhaps, hallucinatory -- if you will pardon my making the inference) logic for the trivialization of the use case -- logic by which we also musn't allow |chapter= together with |chapter_url= (because maybe the editor didn't check that the linked text is true to the original) or |title= together with |trans_title= (because maybe the translated title isn't accurate) or |authorlink= ever (because maybe the linked article is not actually about the author, but rather someone else with the same name). If a new |isbn2= is documented approriately ("For use only where both ISBNs are known to contain identical content at the page-image level e.g. paperback and hardcover editions differing only in their bindings") would we really be opening the door to unprecedented special danger?

If it will help you sleep at night we can call it |isbn2_but_only_if_they_really_have_same_content=, but in truth you are "straining at gnats and swallowing horses", as my Irish great-grandmother used to say. There can be little doubt that many or most ISBNs are not contemporaneously transcribed from the title page of the "work actually consulted" held so sacred here, but rather looked up online, after the fact, once home from the library (or whatever -- you get the idea). Indeed, the scales having now fallen from my eyes about the many potential (if only incipient -- even imagined) instances of insidious inaccuracy indulged by indolent editors, I propose we require that each ISBN be accompanied by a webcam clip of the editor reading aloud the ISBN -- in fact, the entire citation -- straight out of the physical "work consulted", with close-up shot of title page required as well.

The word edition may have been thrown about loosely in this discussion so far, but as has been made abundantly clear we're not talking about distinct editions in the sense of revised content, merely different bindings or printings of identical page content. One frequently sees, especially in titles intended for both library and textbook sales, ISBN J-JJJ-JJJJJJ-J / K-KKK-KKKKKK-K (pbk.) on the title page, which means exactly what one expects it to mean. Of course, there are sometimes minor corrections in the interval between hardcover and softcover issuance, but this happens silently from printing to printing of even the same "edition" (with the same ISBN!) anyway -- if you know what the mysterious runs of numbers like 8 9 10 11 12 13 14 15 16 17 18 19 mean at the bottom of title pages, then you know what I'm talking about. Why aren't you concerned about that?

You say the template "should reflect the usual case as its default". Well, by default |isbn2= will, of course, be null, and that's the usual case.

Finally, your preference that editors "arbitrarily pick one or the other" is, in the context of everything you wrote to that point, contradictory nonsense. If, as you insist, it's hard to know that two ISBNs are interchangeable as the basis of citations, then only one of them is OK -- yet you're suggesting flipping a coin without caring whether you're getting the right one or the wrong one. On the other hand, if picking one or the other arbitrarily is OK, then that must mean you know both of them to be right (i.e. they're identical at the page level), so you may as well help to the reader by including them both.

As to "visual clutter" -- oh, why am I wasting my time? It really makes one think more than twice before bothering to suggest anything. I repeat my earlier comment: It's amazing the amount of verbiage being wasted to characterize as illconceived something that's obviously useful. What a lot of know-it-all bullshit.

EEng (talk) 01:23, 20 March 2013 (UTC)

Here ya go. --— Gadget850 (Ed) ^talk 02:09, 20 March 2013 (UTC)

Markup	Renders as
{{cite book |title=Some Book |id=ISBN 978-0-306-40615-7 (hardback) ISBN 978-0-306-40615-8 (paperback) ISBN 978-0-306-40615-9 (online) ISBN 978-0-306-40615-1 (sign) ISBN 978-0-306-40615-2 (Tijuana bible) ISBN 978-0-306-40615-3 (comic book) ISBN 978-0-306-40615-4 (graphic novel) ISBN 978-0-306-40615-5 (audiobook) ISBN 978-0-306-40615-6 (eBook)}}	Some Book. ISBN 978-0-306-40615-7 (hardback) ISBN 978-0-306-40615-8 (paperback) ISBN 978-0-306-40615-9 (online) ISBN 978-0-306-40615-1 (sign) ISBN 978-0-306-40615-2 (Tijuana bible) ISBN 978-0-306-40615-3 (comic book) ISBN 978-0-306-40615-4 (graphic novel) ISBN 978-0-306-40615-5 (audiobook) ISBN 978-0-306-40615-6 (eBook).

Your example above parodies the use case proposed, while the one below illustrates it realistically. And see my further post, as this same indent level, below. --EEng (talk)

Markup	Renders as
{{cite book |title=The Rise of Statistical Thinking, 1820–1900 | author=Theodore M. Porter| year=1986|publisher= Princeton University Press |id=ISBN 0-691-08416-5 ISBN 0-691-02409-X (pbk)}}	Theodore M. Porter (1986), The Rise of Statistical Thinking, 1820–1900, Princeton University Press, ISBN 0-691-08416-5 (hbk) ISBN 0-691-02409-X (pbk)

Not to seem ungrateful, but why couldn't that have just been the answer to my original query? Were the tortures above some kind of hazing process intended to bestow strength of character through adversity? EEng (talk) 04:53, 20 March 2013 (UTC)

Now can you please explain to me just how a reader is supposed to know which of that long list of ISBNs is the pertinent one, which he should check to verify an assertion, or conversely in your special use-case, how is assured that the editions/versions/printings are equivalent? It is hard enough to get people willing to source-check articles from one specific edition, without asking them to do so in several, let alone having to listen to an audiobook to find out if it is the same as the print version. Protest the runaround all you like, but what you propose represents a massive waste of time for other editors and a reduction of quality for readers.LeadSongDog come howl! 05:30, 20 March 2013 (UTC)

To help you focus, below Gadget's parody example above I've substituted a realistic one. I've also bolded, in earlier posts, the point emphasized ad nauseum from the very beginning, which for good measure I'll repeat here. Get ready. Ready? Here it is:

As has been made abundantly clear we're not talking about distinct editions in the sense of revised content, merely different bindings or printings of identical page content.

Here it is again, this time in bold:

Not distinct editions in the sense of revised content, merely different bindings or printings of identical page content.

One more, time, this time in big type:

Different bindings or printings of identical page content.

I'd be the last to suggest, LeadSongDog, that your username is meant to connote not only possible hallucinations (as previously noted) but prodigious denseness as well -- but can you please get with the program?

-- EEng (talk) 14:57, 20 March 2013 (UTC)

1. There is nothing in policy or common sense requiring the reader to verify that the ISBNs are correct or match each other. Verifiability means that the reader can verify the content of the article. Whether the ISBNs match each other is not part of the content of the article.
2. The point of supplying the ISBN is to aid readers in finding copies of the book. Multiple ISBNs make this easier, not harder, because there is a greater likelihood that at least one of the supplied ISBNs will work to aid the reader to find the book.
3. In practice the reader seeking verification is going to find whatever copy of the book is easiest to find (perhaps using the ISBN to do so), and look in it for the content he or she wishes to verify (ignoring the ISBN or indeed whether it's a correct edition, except possibly if it's a really wrong edition and the desired content isn't there). So all your questions about "how is the reader going to do this?" are completely pointless; none of those things are things that the reader is going to want to do.

—David Eppstein (talk) 05:51, 20 March 2013 (UTC)

In some usual circumstance -- perhaps an obscure foreign book with a confusing title -- I could see supplying an ISBN labeled (?) or something like that, if it's felt that would help someone locate the work in a dusty smoke-filled Karachi bookstore. But in general, any supplied ISBN or ISBNs (whether one, or several) should be one thought to match the page numbers etc. given in the citation. EEng (talk) 14:57, 20 March 2013 (UTC)

I don't see the need for this, but you have an answer. And the standard is to use <s> for strking text, not <del>. And <big> is obsolete. Out. --— Gadget850 (Ed) ^talk 15:06, 20 March 2013 (UTC)

The question previously posed to you, and still not addressed, is why (whether or not you saw the need) the sample markup finally supplied, after exhausting begging, couldn't simply have been given in response to the original request, 30 posts ago? Your fretting about <s> vs. <del> epitomizes the insistence on technical trivia, in lieu of understanding the situation posed, which has characterized this long and wasteful discussion. I repeat: What a lot of know-it-all bullshit. EEng (talk) 16:52, 20 March 2013 (UTC)

Choosing to insult fellow editors rather than work with them is seldom productive. EEng's oft-repeated and now shouted premise that different printings have identical content can for all practical purposes be based only on original research. David evidently doesn't think this matters. While he may be correct, I don't think he is. In any case, that isn't something I need further abuse over. Sorry for trying to answer your question. I'm done. LeadSongDog come howl! 18:12, 20 March 2013 (UTC)

I just picked up three Dover books from my shelf to check. The first one said it was "extensively corrected"; the second said it was unabridged but had a new introduction; the third said it was "unaltered". I don't see how you can construe that as original research, and anyway the Wikipedia original research policy is supposed to be about content, and you are vastly overreaching by trying to apply it to citation data. My feeling is that in the second and third cases it would be appropriate to list both the original and Dover editions. In this case, it shouldn't be done just by giving an ISBN, since the other publication data (e.g. year) is different, but in some cases (electronic vs print isbn for a single book, or library edition vs textbook edition) just listing the multiple isbns in a single citation does make sense. —David Eppstein (talk) 19:26, 20 March 2013 (UTC)

• Verifying an article's citations, page by page, in one ISBN, and then doing that again in a second ISBN, is no more original research than is just doing it in the first one and then stopping.
• Reading, from the title page of a work, the words "This edition published simultaneously as ISBN XXX (hbk) and ISBN YYY (pbk)", and interpreting that appropriately, is no more original research than is getting any other bibliographic information off the title page.

"Abuse" can take many forms, and one of them is to waste others' time and goodwill by stubbornly exhibiting a level of cluelessness so extreme that only the most blunt form of correction has a chance of piercing the armor of self-imposed obliviousness -- so blunt, perhaps, that in another context it might qualify as an insult. But in this context, no. There's no post you've made to this discussion that wouldn't be completely laughable were it not for its contribution to muddling the discussion. EEng (talk) 19:39, 20 March 2013 (UTC)

Abuse, he says. I'll show you abuse. This is abuse. I try to help, and that's the thanks I get? I don't know why I fucking bother. I could have said "this has been asked many times before, check the archives" and left it at that. In fact, if you do bother to look through the archives here, or at many of the talk pages for the Citation Style 1 templates, or those of WP:VPT, you can see if I don't go out of my way to help. And don't shout at us to get your point across. It won't work. --Redrose64 (talk) 22:51, 20 March 2013 (UTC)

Well, since careful reason, examples from experience, detailed explanations, and conscientious response to concerns wasn't working, I thought I'd give it a try. Desperate times call for desperate measures. Look, I'm sure you meant to be helpful, but -- please be candid with yourself -- do you really think a reminder that pp. means pages (plural) was what was needed at that point, especially immediately following my post restating that nothing so far had answered my question? EEng (talk) 02:22, 21 March 2013 (UTC)

An adequate answer was provided by Moxie right at the start; the rest has been unneedful bickering. (And shouting.) Time to put a lid on this discussion? ~ J. Johnson (JJ) (talk) 21:46, 23 March 2013 (UTC)

You refer to the answer here, which was immediately followed by the response

That's what I was doing -- was hoping for something more elegant. But thanks.

So apparently not adequate, though well-intended and appreciated. The actual answer here was provided only much later after much begging. Everything after that was continued counterfactual insistence that this perfectly sensible request was somehow bizarre and wrongheaded.

But I do agree it's time to stop. Those who don't get it yet, probably never will.

EEng (talk) 14:44, 24 March 2013 (UTC)

My two penn'orth - I came across a work today, an encyclopedia in three volumes, each with its own isbn. It was also possible to have all three as a boxed set, no isbn. This was for a general bibliography, not a pinpoint citation, but basically I had to cite all three volumes separately, although it would have been more elegant to have just one citation with three isbns. Must try that isbn-outside-the-cite-book trick.

John of Cromer in Philippines (talk) mytime= Tue 14:33, wikitime= 06:34, 26 March 2013 (UTC)

Error conditions

As far as I know, {{cite web}} is the only citation template that requires title= not be blank, e.g.

{{cite web | title= | url=http://www.foo.com/ | first=James | Last = Ford }} = Ford, James. http://www.foo.com/. {{cite web}}: Missing or empty |title= (help)

{{cite news | title= | url=http://www.foo.com/ | first=James | Last = Ford }} = Ford, James. http://www.foo.com/. {{cite news}}: Missing or empty |title= (help)

while at the same time cite web, for some reason, allows the URL to be omitted:

{{cite web | title=Bob | url= | first=James | Last = Ford }} = Ford, James. "Bob". {{cite web}}: Missing or empty |url= (help)

{{cite news | title=Bob | url= | first=James | Last = Ford }} = Ford, James. "Bob".

This situation makes no particular sense to me. Shouldn't something called "cite web" require a URL? Also, isn't the lack of a title generally a problematic omission in all cases? For example, using {{cite book}} without naming the book seems silly.

In working on the Lua migration of the citation templates, I'm wondering what to do about the error conditions. We can, of course, enforce a rule that {{cite web}} must have a title but not the other templates, if that is what's needed. But that seems rather arbitrary. Do people here have suggestions for what error checks you would like to see? Dragons flight (talk) 04:57, 17 March 2013 (UTC)

I find it sensible that when a URL is produced, there should be a forced title for all of the templates. cite web might be the special case where the title is expected to be an online one (not sure that's relevant).

Forcing a URL though? Not sure about that. Even if it is cite web. --Izno (talk) 13:39, 17 March 2013 (UTC)

Cite encylopedia - Lua

{{Cite encyclopedia}} now uses the Lua version. Please report any issues. --— Gadget850 (Ed) ^talk 14:48, 17 March 2013 (UTC)

{{Cite Nuttall}}, which relies on this template, is broken (editorlink param). Bob Burkhardt (talk) 22:08, 17 March 2013 (UTC)

I fixed it, but it was never a Lua issue. That template was broken due to an erroneous edit last June. Dragons flight (talk) 22:26, 17 March 2013 (UTC)

Thank you. Bob Burkhardt (talk) 01:49, 18 March 2013 (UTC)

the replacement of the modular {{cite encyclopedia}} application/{{citation/core}} backend with the present iteration of the monolithic citation/cs citation/cs1 script is a mixed bag.

i didn't test enough to see if there are coding bugs (this should be done by the developers prior to deployment. syntax flaws are UNACCEPTABLE). several glaring logic bugs (the nonsensical dependencies among parameters) have been happily removed. some design bugs, namely the equally nonsensical unavailability of the full parameter set to the citation class have also been fixed. unfortunately other, equally glaring and potentially show-stopping design bugs remain. the most serious is the fact that the porting retains the parameter ambiguity of the previous system when it comes to args.title. this may mean either (a) title a work fragment (eg periodicals) or (b) title of the whole work (eg book). this also unacceptable ambiguity leads to uneccessary and inefficient workarounds of the args.booktitle kind.

as with all bugs it can lead to unexpected results such as below:

Markup	Renders as
*{{cite encyclopedia|ref=harv|last=Author|date=2013|editor-last=Editor|article=Article|encyclopedia=Encyclopedia|type=Type}}	Author (2013). "Article". In Editor (ed.). Encyclopedia (Type). {{cite encyclopedia}}: |last= has generic name (help); Invalid |ref=harv (help)

cf. {{cite book}}:

Markup	Renders as
*{{cite book|ref=harv|last=Author|date=2013|editor-last=Editor|contribution=Article|title=Encyclopedia|type=Type}}	Author (2013). "Article". In Editor (ed.). Encyclopedia (Type). {{cite book}}: |last= has generic name (help); Invalid |ref=harv (help)

as you see, the positioning and punctuation of type in the encyclopedia citation is wrong. this is because type is coded to come after (work) title. but in the encyclopedia/periodical case it instead appears after (fragment) title. the software has no routine to distinguish between the two. such a routine shouldn't be written; instead the parameters should be renamed and the relevant doc be rewritten so there is no semantic ambiguity.

my current evaluation of the lua-based citation implementation is neutral. 70.19.122.39 (talk) 12:43, 18 March 2013 (UTC)

Re title: I think you are referring to the title issue described in Module_talk:Citation/CS1#Weird_link_formatting_for_encyclopedia_class this discussion, where the old version used title in an odd manner that caused odd output. This has been fixed. I have reported the issue with type (which was not supported in the old version). I will work on the documentation. --— Gadget850 (Ed) ^talk 14:16, 18 March 2013 (UTC)

i'm referring to the fact that title may mean either a work title (args.BookTitle=args.Title) or a fragment title (see section "-- Account for the oddity that is {{cite conference}}, before generation of COinS data."). this messes type up, as evident in the local tcommon routines of Module:Citation/CS1. 70.19.122.39 (talk) 14:36, 18 March 2013 (UTC)

Yes, "title" has an ambiguous meaning that usually translates as "article title", but sometimes translates as "encyclopedia title" or similar. We can't standardize that without breaking the existing base of citations, so it won't be fixed. Regarding "type", it has never been on the list of parameters expected by "cite encyclopedia", so it isn't something we test against (same with "series" and "issue" and "agency", and other options that aren't even supposed to be used for cite encyclopedias). We don't currently check whether the user submitted an invalid parameter set with extra options, though we might in the future. Dragons flight (talk) 15:08, 18 March 2013 (UTC)

P.S. If people want to add additional options, like type=, to classes that don't currently use them then we can work on making sure that they will look right; however, it would be up to users to define what parameters they want to support. Dragons flight (talk) 15:12, 18 March 2013 (UTC)

nope, this is wrong. it is much more important to fix meaning ambiguous to humans than to guarantee compatibility for software. first, find a way to correct the ambiguity even if it takes all your time and resources. THEN you can proceed with deployment of the new system without the old bugs. this is a unique opportunity to make the citation system understandable and logical. your way consigns a vitally important function of wikipedia to continuing confusion.

what set of parameters had been "expected" by {{cite encyclopedia}} is irrelevant. expected by whom? and when? if adding the full set of params to citation classes will not unduly break performance, then they should be there, period. and the code should properly present any and all parameters to readers. let the editors decide what to use, don't dictate to them. 70.19.122.39 (talk) 15:42, 18 March 2013 (UTC)

"Title" is accepted as an ambiguous word in the printed style manuals that inspired the cite templates. For example, Chicago Manual of Style 16th ed. p. 661 states "titles of larger works (e.g., books and journals) are italicized; and titles of smaller works (e.g., chapters, articles) are presented in roman and enclosed in quotation marks." If you want to change the parameter names now, it is up to you to think of a way to automatically change all the existing instances of the templates in the encyclopedia, and change all the software that helps people create citation templates. If you can't do that, or you can't find a way to live with the ambiguity, you will have to abandon your project. Jc3s5h (talk) 16:29, 18 March 2013 (UTC)

Lets look at what is going on here with title:

The old version of cite encyclopedia included:

|Title={{{encyclopedia|{{{title|}}}}}}
|TransTitle={{{trans_chapter|}}}
|TransItalic={{{trans_title|}}}
|IncludedWorkTitle={{{title|{{{article|}}}}}}

So, if you defined title but not encyclopedia then title would be displayed twice:

Cite encyclopedia comparison

Wikitext	{{cite encyclopedia|title=Europe|url=http://encarta.msn.com/encyclopaedia_761570768/Europe.html}}
Live	Europe.
Sandbox	Europe.

The old method is just plain wrong. I discussed this before with little response and let it go. I should have fixed it then and there and we would not be here. I think we should treat title as the quote marked included title and fix uses as they come up. I also think we should do the same with the old template, which may be reused on other wikis. --— Gadget850 (Ed) ^talk 16:52, 18 March 2013 (UTC)

Gadget850, are you suggesting that within the "cite encyclopedia" template, the parameter "title" means the title of an article, but within other types of cite template, such as "cite book", it could mean something else? Because it seemed to me that 70.19.122.39 wanted "title" to always mean the same thing for all citation style 1 templates.

But it already doesn't always mean the same thing, and I think it's too late to change. E.g. in {{cite book}} it's the title of a book while in {{cite conference}} it's the title of an article within a book (and the book title is given by the |booktitle= parameter). —David Eppstein (talk) 17:30, 18 March 2013 (UTC)

and in {{cite news}} or {{cite journal}} it's the title of the article within the work. --Redrose64 (talk) 17:58, 18 March 2013 (UTC)

(edit conflict)Yes, title varies among templates. The problem is that these templates were developed separately and then updated to use core. When I updated cite conference, I had to use the parameters that were already in use. For encyclopedia, it should mean the article within the encyclopedia and nothing else. --— Gadget850 (Ed) ^talk 18:11, 18 March 2013 (UTC)

the limitations of the old setup are known. just see the logical + linguistic hoops in the posts above. if in {{cite conference}} title means "paper title" why not rename the param paper? it is unambiguous and exactly describes the cited item. similarly in {{cite news}} etc. why not just use article instead of title? an article is what is being cited, and i challenge anybody to prove that editors will understand article to mean something other that "article title". this is/was a unique opportunity to design a citation system that is logical and accessible from the ground up. as well as establish an ancillary manual of style (the doc) for editors and readers that is simple and readily understandable by non-experts, so let's forget the expert-based systems for the moment. a parallel citation system based on lua or whatever could be designed correctly to coexist with the present, and editors should be given guidance to use it for new citations, while (with some automation) the older style is gradually replaced. or, and to continue with title in a very rough example: define eg book as an additional alias for work/title in {{citation/core}} and {{cite book}} (<- the main offender). replace all instances of title in the {{cite book}} code with book. automate the replacement of |title= with |book= in article space {{cite book}} citations. then at least you will be left with title unambiguously meaning title of fragment-of-work. and even that should be phased out because we don't cite "titles". we cite works that may be articles, books, films, signs, etc. 70.19.122.39 (talk) 13:38, 19 March 2013 (UTC)

If we were starting from scratch, I would write a specification to work from. But, we have 23 base templates that were independently created. They have been kludged to fit the CS1 style and work pretty well considering.

3Es: engineering, education, enforcement. {{Cite conference}} has been mentioned. It uses title for the included work and booktitle for the main work. It has just under 5000 uses. We could update the parameters, make booktitle throw an error message and use a bot to fix all the uses. But trying to educate all the editors in the new use is going to involve pain and heartache. Or; we could just keep it as is and use the Lua technology to make it work transparently. --— Gadget850 (Ed) ^talk 14:26, 19 March 2013 (UTC)

i too, enjoy a good joke, but this is not very funny. in this very thread i show how handling of at least one error by the present lua implementation is anything but transparent. now whether the "klutzes" that make up the markup-based CS1 performed well is surely a matter of opinion. discussions in several talk pages about problem after problem (real or perceived) seem to suggest otherwise. the point is that the lua implementation is starting from scratch. it completely does away with the previous setup, using (at this point) templates only as wrappers for calling modules and passing them parameters+data. it could (should) have been designed intelligently to do away with all the bugs plaguing the previous system whether these are bugs in code, logic, process, or design. but right now the implementation, and comments on this thread, show that this is not the case, and it may be hopeless. so much for "lua technology" (?) fixing the citation system, solving the eurozone crisis, and brewing everybody a nice cup of tea. as a sidenote, it is interesting how the first reaction of yourself and others is finding an explanation to deny change requests by editors, instead of finding a way to actually implement them. and most of the editor requests imo are legitimate: that is why many are submitted again and again by different editors through weeks, months, and years. the explanations on the other hand, seem trivial: (a) well, that's how it's been done... (b) some irrelevant manual says so (as if we're publishing for a university press, specialist house, or the new york times) (c) can't do it because some editor might disagree about some vaguely defined item and then what are we going to do... (d) let's have an rfc! no, make that a hundred (e) there have been lots and lots of discussions that had no conclusion, so there (this actually is a nice example of recursion or self-reference, as most of the negative comments in these discussions are by those who then bring this up as an argument) (f) i have a toothache (ok the last is probably original research). fun in la-la land. 70.19.122.39 (talk) 00:18, 21 March 2013 (UTC)

You seem to misunderstand. The Lua implementation of citations is not starting from scratch. One of the design requirements is to fully support the existing parameter set so that existing templates can be migrated to Lua without losing the content of the preexisting citations throughout Wikipedia. That means working around any number of stupid historical choices for parameter naming and function. We can still fix many formatting and logic bugs (and we are working on that; it is still early), but we need to support all of the historical use cases which limits the set of changes that are possible. It was never the plan to make Lua citations completely independent of the currently existing template scheme. While there could be some value in reengineering everything from scratch, and rethinking parameter specifications and such, that is not what we are doing here. Dragons flight (talk) 00:36, 21 March 2013 (UTC)

This is one of the reasons I prefer {{citation}}: one set of params to remember instead of 23 subtly-different sets. —David Eppstein (talk) 02:38, 20 March 2013 (UTC)

Page enumeration conflicts

According to the CS1 standard, the options the page=, pages=, and at= can all be used to provide page information, but no more than one of these is ever displayed, with page= overriding pages= and both overriding at=. As part of the Lua migration, we have added the hidden category Category:References with multiple page specifications to any page where a reference tries to simultaneously use more than one of these options. This will allow these conflicting parameters sets to be identified and cleaned up. Dragons flight (talk) 17:49, 19 March 2013 (UTC)

I think it would be better to throw a red error message in such cases, rather than just using a category, that way, the editor adding the superfluous detail is alerted to it. Otherwise, a subsequent editor nay not know which of the values given is correct. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:13, 20 March 2013 (UTC)

I had been considering that as well. I think that both page and pages are added when the editor is including the number of pages in the work, which we don't need. --— Gadget850 (Ed) ^talk 15:19, 20 March 2013 (UTC)

In the future I would agree with you, but not during the Lua migration. To simply turn on an error message that is likely to hit some few thousand pages (once all citation modes are migrated) seems unnecessarily disruptive. If after the migration the backlog gets cleared, then I think it would make sense to convert this to a visible error. Dragons flight (talk) 15:25, 20 March 2013 (UTC)

I concur, though not for the reasons Editor Andy Mabbett suggests. I've been plodding through the articles listed at Category:Articles with incorrect citation syntax. Finding the broken citations is relatively easy when there is a red error message. You'd think that editors would see the red error text and fix their citations. Apparently not.

—Trappist the monk (talk) 15:33, 20 March 2013 (UTC)

How to Cite eBooks

I have issues citing eBooks, in particular when referencing a page, which may vary by varying the font size. In Kindle the exact location can be expressed with a number called Location.

So maybe some rules and/or parameters are necessary, Or not? Carlotm (talk) 11:44, 20 March 2013 (UTC)

See WP:Page numbers. You can use |at= with the CS1 templates. --— Gadget850 (Ed) ^talk 12:23, 20 March 2013 (UTC)

Thanks Gadget850. So I suppose a citation like the following one would be fine.

Epstein, Catherine (2010). "Model Nazi: Arthur Greiser and the Occupation of Western Poland". New York: Oxford University Press. location 2772-2778 (Kindle Edition). — Preceding unsigned comment added by Carlotm (talk • contribs) 07:54, 21 March 2013 (UTC)

Using CS1:

Markup	Renders as
{{cite book |last=Epstein |first=Catherine |year=2010 |title=Model Nazi: Arthur Greiser and the Occupation of Western Poland |location=New York |publisher=Oxford University Press |at=location 2772-2778 (Kindle)}}	Epstein, Catherine (2010). Model Nazi: Arthur Greiser and the Occupation of Western Poland. New York: Oxford University Press. location 2772-2778 (Kindle).

--— Gadget850 (Ed) ^talk 10:06, 21 March 2013 (UTC)

Cite web format and language parameter position in rendered citation

Markup	Renders as
{{cite web |url=http://www.bnp.gob.pe/snb/data/periodico_mural/2009/7/fechas_civicas/28dejulio.pdf |title=Aniversario de la Proclamacion de la Independencia del Perú |language=Spanish |format=pdf |accessdate=2013-03-21 }}	"Aniversario de la Proclamacion de la Independencia del Perú" (pdf) (in Spanish). Retrieved 2013-03-21.

The format text normally follows the pdf icon when |language= isn't used. It seems odd to insert the language between the icon and the file format so perhaps these two parameters should be swapped in {{citation/core}}. Right now, the code looks like this:

{{
  #if: {{{language|}}}
  | (in {{{language}}})
}}{{
  #if: {{{format|}}}
  | ({{{format}}})
}}

—Trappist the monk (talk) 13:48, 21 March 2013 (UTC)

That is fixed in the Lua version. Not sure when web will be deployed:

Cite web comparison

Wikitext	{{cite web|accessdate=2013-03-21|format=pdf|language=Spanish|title=Aniversario de la Proclamacion de la Independencia del Perú|url=http://www.bnp.gob.pe/snb/data/periodico_mural/2009/7/fechas_civicas/28dejulio.pdf}}
Live	"Aniversario de la Proclamacion de la Independencia del Perú" (pdf) (in Spanish). Retrieved 2013-03-21.
Sandbox	"Aniversario de la Proclamacion de la Independencia del Perú" (pdf) (in Spanish). Retrieved 2013-03-21.

--— Gadget850 (Ed) ^talk 14:01, 21 March 2013 (UTC)

 Fixed --— Gadget850 (Ed) ^talk 19:35, 29 March 2013 (UTC)

Cite pmid

edited to finish what I started; guess I must have inadvertantly clicked save

I just replaced this citation a {{cite pmid|23019641}} with a {{cite journal}} equivalent at R. Duncan Luce. This page was listed at the end of all of the Template: pages at Category:Articles with incorrect citation syntax. None of the usual error messages were displayed and none of the other citations seemed to have anything wrong with them so I took a chance and replaced the {{cite pmid}}.

So I guess my questions are: how did {{cite pmid}} cause this error? Why wasn't there an error message? Why was the R. Duncan Luce page listed with the Templates?

—Trappist the monk (talk) 14:31, 21 March 2013 (UTC)

It redirects to {{Cite doi/10.1126.2Fscience.1229851}}, but I don't see anything there. --— Gadget850 (Ed) ^talk 14:32, 21 March 2013 (UTC)

Edit request

Admins, the template {{Cite news}} got messed up. See this. Please correct it. Nataev (talk) 14:28, 21 March 2013 (UTC)

Don't use a language template in the citation template, use the |language= parameter. --— Gadget850 (Ed) ^talk 14:37, 21 March 2013 (UTC)

OK, fair enough! Thanks! Nataev (talk) 18:03, 21 March 2013 (UTC)

 Fixed --— Gadget850 (Ed) ^talk 19:34, 29 March 2013 (UTC)

lastauthoramp parameter

Unlike other "cite" templates and the {{citation}} template, {{cite journal}} does not respect |lastauthoramp=yes. For example:

• {{cite journal |last=Last1 |last2=Last2 |year=2000 |title=Title |journal=Journal |lastauthoramp=yes }} → Last1; Last2 (2000). "Title". Journal. {{cite journal}}: Unknown parameter |lastauthoramp= ignored (|name-list-style= suggested) (help)CS1 maint: numeric names: authors list (link)
• {{citation |last=Last1 |last2=Last2 |year=2000 |title=Title |journal=Journal |lastauthoramp=yes }} → Last1; Last2 (2000), "Title", Journal {{citation}}: Unknown parameter |lastauthoramp= ignored (|name-list-style= suggested) (help)CS1 maint: numeric names: authors list (link)
• {{cite book |last=Last1 |last2=Last2 |year=2000 |title=Title |lastauthoramp=yes }} → Last1; Last2 (2000). Title. {{cite book}}: Unknown parameter |lastauthoramp= ignored (|name-list-style= suggested) (help)CS1 maint: numeric names: authors list (link)

Could this be fixed please? Peter coxhead (talk) 16:59, 24 March 2013 (UTC)

This will be fixed in the next release of Module:Citation/CS1. Personally though, I think that's a rather silly parameter. It would make more sense to either always end with an ampersand, or never do so. I don't really see the point of having it as a user configurable parameter (which is turned on for fewer than 1 in 100 citations). Dragons flight (talk) 17:33, 24 March 2013 (UTC)

Well, I personally prefer the ampersand, but I don't feel very strongly about it. Peter coxhead (talk) 21:25, 24 March 2013 (UTC)

I just reported two display issues at Module talk:Citation/CS1. --— Gadget850 (Ed) ^talk 18:23, 24 March 2013 (UTC)

 Fixed --— Gadget850 (Ed) ^talk 19:33, 29 March 2013 (UTC)

lua: author masking

cf. {{cite web}} (non-lua)

70.19.122.39 (talk) 12:32, 25 March 2013 (UTC)

Lua is using author-mask but not the alias authormask. Reported. --— Gadget850 (Ed) ^talk 12:50, 25 March 2013 (UTC)

thank you. will bulk-edit my cites accordingly. 70.19.122.39 (talk) 13:01, 25 March 2013 (UTC)

 Fixed --— Gadget850 (Ed) ^talk 10:56, 29 March 2013 (UTC)

lua: |type= in {{cite news}}

Markup	Renders as
{{cite news|author=Author|title=Article|newspaper=Newspaper|type=Type}}	Author. "Article". Newspaper (Type). {{cite news}}: |author= has generic name (help)

cf. {{cite encyclopedia}} (lua, fixed)

Markup	Renders as
{{cite encyclopedia|author=Author|article=Article|encyclopedia=Encyclopedia|type=Type}}	Author. "Article". Encyclopedia (Type). {{cite encyclopedia}}: |author= has generic name (help)

70.19.122.39 (talk) 12:59, 25 March 2013 (UTC)

 Fixed --— Gadget850 (Ed) ^talk 10:57, 29 March 2013 (UTC)

lua: |format= in {{cite journal}}

70.19.122.39 (talk) 13:06, 25 March 2013 (UTC)

format is a descriptor for url; it really should be suppressed if url is undefined:

Cite journal comparison

Wikitext	{{cite journal|author=Author|journal=Journal|title=Article|type=Type}}
Live	Author. "Article". Journal (Type). {{cite journal}}: |author= has generic name (help)
Sandbox	Author. "Article". Journal (Type). {{cite journal}}: |author= has generic name (help)

But it does work properly with url:

Cite journal comparison

Wikitext	{{cite journal|author=Author|format=Format|journal=Journal|title=Article|type=Type|url=http://example.org}}
Live	Author. "Article" (Format). Journal (Type). {{cite journal}}: |author= has generic name (help)
Sandbox	Author. "Article" (Format). Journal (Type). {{cite journal}}: |author= has generic name (help)

--— Gadget850 (Ed) ^talk 13:18, 25 March 2013 (UTC)

no it's not, and it shouldn't. it is a parameter that gives further info depending on media type: in "cite book" it normally (though not exclusively) refers to binding. in "cite web" or digital media may refer to file format (eg an e-book could be in EPUB format, a downloadable journal may be in Kindle format etc. no urls may be provided in these cases.) 70.19.122.39 (talk) 13:28, 25 March 2013 (UTC)

You are confusing format and type. Per {{cite book}} and {{cite web}}:

• format: Format of the work referred to by url; examples: PDF, DOC, XLS
• type: Provides additional information about the media type of the source

--— Gadget850 (Ed) ^talk 13:42, 25 March 2013 (UTC)

and thus ambiguity in parameter meaning is unhappily retained. i am not discussing "type". "format" in book citations means primarily binding format: paperback, cloth etc. "format" in url/digital media citations means primarily (electronic/virtual) binding format: mobipocket format, portable document format etc. see? simple. so format can and SHOULD mean the same across the cite system. let's make it so. 70.19.122.39 (talk) 14:01, 25 March 2013 (UTC)

The format parameter is consistent across all 23 templates. --— Gadget850 (Ed) ^talk 14:29, 25 March 2013 (UTC)

please stop, this is not making sense.

1. spot the "consistency" in the display bug below:

Markup	Renders as
{{cite book|author=Author|date=Date|chapter=Chapter|work=Book|format=Format|type=Type}} vs. {{cite journal|author=Author|date=Date|title=Article|journal=Journal|format=Format|type=Type}}	Author (Date). "Chapter". (Type). {{cite book}}: |author= has generic name (help); |format= requires |url= (help); |work= ignored (help); Check date values in: |date= (help); Missing or empty |title= (help) vs. Author (Date). "Article". Journal (Type). {{cite journal}}: |author= has generic name (help); |format= requires |url= (help); Check date values in: |date= (help)

2. the dependency of |format= on |url= is just wrong and should be undone. the doc should immediately reflect this.

in citations, "format" has a more far-ranging meaning than the very specific "url". consequently, it can justifiably be used by editors to convey info about the item cited. assuming the citation is properly structured the average reader can easily divine the meaning of "format" from the context:

• Author (paperback) Book (print) - in this case the item (book) type/medium is print and is formatted as paperback
• Performer (mono) MusicWork (CD) - in this case the item (audio) type/medium is a CD and the audio format is monophonic
• Author (CD) Book (audio) - in this case the item (book) type/medium is audiobook and the format is CD
• Author (EPUB) Book (digital) - in this case the item (book) type/medium is digital/e-book and the format is EPUB
• Creator (headstone) Sign (marble) - in this case the item (sign/visual) type/medium is marble and the work's format is headstone

"format" is a variable that can assume many more values than "url" which only refers to an internet address. therefore it is a more significant variable.

in the context of designed dependencies a basic principle of sound design is to make less significant variables dependent on more significant ones, not the other way around.

it is truly amazing that i have to expend time and effort to explain these self-evident things. please, just fix it or at least stop making condescending excuses. 70.19.122.39 (talk) 12:55, 26 March 2013 (UTC)

Looks like you're having much the same problem I ran into recently. See #verbiage EEng (talk) 19:05, 27 March 2013 (UTC)

well what can i say. i appreciate the effort, but responsiveness needs improvement. 70.19.122.39 (talk) 01:11, 28 March 2013 (UTC)

It appears that your vision of format conflicts with the manner in which it has been documented since the original templates were implemented. If you use a field in a manner in which it was not intended, then we cannot anticipate that changes affect your use or misuse. Each field in each template is well documented and is fairly consistent across the series. format is implemented and documented exactly the same across all templates: it reflects the format of the web link, not the medium type of the source which is intended for type. --— Gadget850 (Ed) ^talk 15:52, 26 March 2013 (UTC)

i have no "vision" for |format=. i explained above how "format" may be used logically and simply to make a citation more accessible to readers. making a citation more accessible is making it more easy to verify, and that is the only reason a citation exists, or else might as well do away with sources altogether. the illogical format/url dependency is a relatively recent change. previously the documentation assigned |format= a more pertinent and correct place. when the doc is fixed (it is substandard) you may want to restore it. anyway, don't want to be discussing this forever, you either see it or you don't. 70.19.122.39 (talk) 13:55, 27 March 2013 (UTC)

At this point I am not going to convince you of anything. The documentation will continue to be substandard until you either boldly change it or become involved in discussing specific issues. I am quite tired of the neeping and nopping. --— Gadget850 (Ed) ^talk 19:16, 27 March 2013 (UTC)

i have gone into details several times before. i have also mentioned that you imo cannot treat doc for wildly different sources as a monolithic item. technically the citations share certain parameters and certain display conventions (and code). this is of no importance to users, and the technical doc is better suited to developers. editors need to know how best to cite a specific type of source, and readers need to understand what the citation describes. so it is that "vision" thing again. don't think about parameters, think about what any one parameter means to someone (reader/editor) who hasn't seen that particular type of citation before. i thought you welcomed reports of problems. i see this as a problem. 70.19.122.39 (talk) 01:11, 28 March 2013 (UTC)

btw, unless the tfd process becomes more rational, don't expect any involvement on my part. i'm not gonna do work that can be deleted so easily, and with minimal discussion. 70.19.122.39 (talk) 01:11, 28 March 2013 (UTC)

At this edit on 30 May 2011, {{citation}} got text describing the purpose of |format=. The new {{cite book}} documentation of 9 September 2006 has text describing the purpose of |format= from its inception. In both cases, |format= is grouped under |url= implicitly identifying it as pertaining to |url=.

I have found no uproar in the {{citation}} and {{cite book}} talk pages to suggest that the meaning of |format= is or should be other than to identify the format of the resource addressed by |url=. You wrote: previously the documentation assigned |format= a more pertinent and correct place. Can you point me to that previous documentation?

—Trappist the monk (talk) 20:05, 27 March 2013 (UTC)

i will try to find it. however it may have been in the doc for Wikipedia:Citing sources or a similar, equally high profile page, and i will check. for years, i kept seeing "format" in {{cite book}} citations with such info as "hardcover" etc. nobody complained about that either. 70.19.122.39 (talk) 01:11, 28 March 2013 (UTC)

well i checked all the pages i thought possibly would carry the info but i could not find it. all i can say is that i do recall "format" being explained in the manner i suggested, and a "cite book" example being given with "format" describing the (physical) binding. i still stand by the merit of my suggestion. in any case, if there is a url involved, for verification purposes, the editor is no longer are citing a (physical) book, but a digital medium, and "type" as well as "format" should reflect that. the editor may have consulted a physical book when citing, but that is irrelevant: you are telling readers to verify the cited information online. edit: i assume the online publisher if different, (should get equal billing in |publisher=) is reliable 70.19.122.39 (talk) 14:35, 29 March 2013 (UTC)

lua: |department=

{{cite news}}:

Markup	Renders as
{{cite news|author=Author|title=Article|department=Department|newspaper=Newspaper}}	Author. "Article". Department. Newspaper. {{cite news}}: |author= has generic name (help)

{{cite journal}}:

Markup	Renders as
{{cite journal|author=Author|title=Article|department=Department|journal=Journal}}	Author. "Article". Department. Journal. {{cite journal}}: |author= has generic name (help)

70.19.122.39 (talk) 13:23, 25 March 2013 (UTC)

Comfirmed. --— Gadget850 (Ed) ^talk 13:46, 25 March 2013 (UTC)

Cite journal comparison

Wikitext	{{cite journal|department=Bryan on Scouting|first=Bryan|journal=Scouting|last=Wendell|title=Calendar of New Merit Badges|url=http://blog.scoutingmagazine.org/merit-badge-calendar/}}
Live	Wendell, Bryan. "Calendar of New Merit Badges". Bryan on Scouting. Scouting.
Sandbox	Wendell, Bryan. "Calendar of New Merit Badges". Bryan on Scouting. Scouting.

 Fixed --— Gadget850 (Ed) ^talk 12:27, 29 March 2013 (UTC)

lua: en dash (multiple entries) in |volume=

separated by en dash (per MOS), not bolded

Markup	Renders as
{{cite book|title=Title|series=Series|volume=1–2}}	Title. Series. Vol. 1–2.

cf. separated by hyphen (incorrect), bolded

Markup	Renders as
{{cite book|title=Title|series=Series|volume=1-2}}	Title. Series. Vol. 1–2.

70.19.122.39 (talk) 14:12, 25 March 2013 (UTC)

That is odd. By design, five or more characters are no longer bolded, but that is only three:

Cite book comparison

Wikitext	{{cite book|sandbox=true|series=Series|title=Title|volume=1–2}}
Live	Title. Series. Vol. 1–2. {{cite book}}: Unknown parameter |sandbox= ignored (help)
Sandbox	Title. Series. Vol. 1–2. {{cite book}}: Unknown parameter |sandbox= ignored (help)

Cite book comparison

Wikitext	{{cite book|sandbox=true|series=Series|title=Title|volume=1-2}}
Live	Title. Series. Vol. 1–2. {{cite book}}: Unknown parameter |sandbox= ignored (help)
Sandbox	Title. Series. Vol. 1–2. {{cite book}}: Unknown parameter |sandbox= ignored (help)

Cite book comparison

Wikitext	{{cite book|sandbox=true|series=Series|title=Title|volume=1234}}
Live	Title. Series. Vol. 1234. {{cite book}}: Unknown parameter |sandbox= ignored (help)
Sandbox	Title. Series. Vol. 1234. {{cite book}}: Unknown parameter |sandbox= ignored (help)

Cite book comparison

Wikitext	{{cite book|sandbox=true|series=Series|title=Title|volume=12345}}
Live	Title. Series. Vol. 12345. {{cite book}}: Unknown parameter |sandbox= ignored (help)
Sandbox	Title. Series. Vol. 12345. {{cite book}}: Unknown parameter |sandbox= ignored (help)

--— Gadget850 (Ed) ^talk 14:26, 25 March 2013 (UTC)

Hyphen to endash conversion has been added for the short volume case. The bolding is intentional and related to style consistency across cite book / cite journal / etc. Dragons flight (talk) 15:05, 25 March 2013 (UTC)

Oh, I see, you are complaining that "1–2" in particular wasn't rendered as bold. Dragons flight (talk) 15:25, 25 March 2013 (UTC)

Fixed in sandbox. Dragons flight (talk) 15:27, 25 March 2013 (UTC)

please tell me. is there any reason for this?

 if ( Volume ~= nil and Volume ~="" )
    then
        if ( mw.ustring.len(Volume) > 4 )
          then Volume = sepc .." " .. Volume
          else Volume = " <b>" .. hyphentodash(Volume) .. "</b>"
        end
    else Volume = "" end

so "volume data" are bolded when they are 4-characters-long or less

but "volume data" are not bolded otherwise.

congratulations! you introduced one more design bug (in the form of presentational ambiguity)

what on earth? why not leave "volume data" always bolded? or always in plaintext? is there a competition to cram as many "if" statements (process forks) as possible? if you want to have volume numbers bolded and volume titles in plaintext why not just introduce different parameters? but I can hardly think of a need for that.

am I missing some other unfathomable design purpose here? 70.19.122.39 (talk) 13:13, 26 March 2013 (UTC)

The "purpose" is that people apparently like the bolding unless the volume label was long. For example:

People seem to like bolding on short numbers:

{{cite book | first = John | last = Doe | title=History of America | volume=1 | date=1978 | publisher=McMillan and Company | location=New York }}

• Doe, John (1978). History of America. 1. New York: McMillan and Company. 

But people also complain about:

{{cite book | first = John | last = Doe | title=History of America | volume=American Revolution to Civil War | date = 1978 | publisher=McMillan and Company | location=New York }}

• Doe, John (1978). History of America. American Revolution to Civil War. New York: McMillan and Company. 

So, now we have:

• Doe, John (1978). History of America. Vol. 1. New York: McMillan and Company.
• Doe, John (1978). History of America. Vol. American Revolution to Civil War. New York: McMillan and Company.

If you can convince the Wikipedia editor community to settle on using one style or the other consistently, then all the better. However, if the community requests content-dependent formatting then that is what they are going to get. For the record, I'm just reporting the logic as it was communicated to me. The actual decision to implement this approach was made before I got involved, so I'm not sure where the original discussions of this may have taken place. Dragons flight (talk) 14:58, 26 March 2013 (UTC)

There have been multiple discussions about volume bold on the talk pages for Help:CS1, {{Citation}} and {{citation/core}}. At that point, we did not really have the ability to selectively bold the volume, so those discussions went nowhere. --— Gadget850 (Ed) ^talk 15:03, 26 March 2013 (UTC)

i don't need to convince anybody: volume data was always bolded by default before, and i am unaware of any convincing being done then. it is also not true that we could not "selectively bold the volume". all you had to do was add ''' markup around the data to unbold it. i did it in 1 instance. the doc was mum on that workaround so i suppose whoever needed convincing was convinced, or else thankfully the powers that be were unaware of the workaround, and thus could not screw things further. unlike in this case.

i shouldn't have to convince you that readers (they are the most significant target, the editors are second) are not well served by more ambiguity. and that is exactly what the new presentation of |volume= introduces, being sometimes bold and sometimes not. it is imo much better for editors (and readers) if you introduce eg the independent parameters "volume" (for volume title) and "number" (for volume number) as i suggested above. and then you can preformat them as you see fit. 70.19.122.39 (talk) 13:58, 27 March 2013 (UTC)

also, consider situations where multiple vols. may be cited:

Markup	Renders as
{{cite encyclopedia|editor=Editor|chapter=Editor Notes|work=Multi-volume Work|Volume=1–3, 6, 18–22}}	Editor (ed.). "Editor Notes". Vol. 1–3, 6, 18–22. {{cite encyclopedia}}: |editor= has generic name (help); |work= ignored (help); Missing or empty |title= (help)

numbers, not bolded. i used {{cite encyclopedia}} because is the recommended template for multi-vol works. also, is it impossible to have a 3-letter volume title? i don't think so. 70.19.122.39 (talk) 14:20, 30 March 2013 (UTC)

i notice that trailing punctuation for |series= has been removed if volume data<5. although this is a good idea, it compounds the confusion, because the punctuation is retained otherwise. see also my related proposal here. 70.19.122.39 (talk) 13:09, 3 April 2013 (UTC)

in related consideration, should "volume title" even be in plain text? every other "title" or "wannabe-title" is preformatted. 70.19.122.39 (talk) 13:22, 3 April 2013 (UTC)

{{Cite book}}

I am not sure what has happened to {{Cite book}} but any URL not in the URL parameter is no longer working. I understand that having a URL not in the URL parameter is not the way to do things but never the less we have thousands and thousands of links set up out side the normal URL parameter. What has happened anyone working on this problem?Moxy (talk) 18:43, 25 March 2013 (UTC)

• Wikipedia:Village pump (technical)#Linking problem to Google books
• Module talk:Citation/CS1#URL in page number

--— Gadget850 (Ed) ^talk 18:45, 25 March 2013 (UTC)

Thank you for that link - Looking at that talk looks like people are not aware that at our content guideline we tell editors to link like this see - WP:BOOKLINKS.Moxy (talk) 18:58, 25 March 2013 (UTC)

To be more specific, it is not any URL. It is a URL stuffed into the page=/pages= field or the volume= field, and then only if the URL contains a hyphen ("-"). I suspect the number of broken URLs is actually pretty small, nonetheless we will fix that part. However, having a URL in the pages= or volume= field is incompatible with the COinS metadata generated by the citation templates (and always has been), so we should not be encouraging people to do that. Dragons flight (talk) 19:04, 25 March 2013 (UTC)

 Fixed I added a note to BOOKLINKS, but if the practice is already in use, it will continue. --— Gadget850 (Ed) ^talk 19:30, 25 March 2013 (UTC)

ISBN

I usually change any dash in isbn to {{nbhyph}}, because Sod's law says the line will break there. However now although {{nbhyph}} gets transformed correctly to &#‍x2011; that is how it stays. So isbn gets shown as for instance 978&#‍x2011;1892214973

John of Cromer in Philippines (talk) mytime= Tue 13:53, wikitime= 05:53, 26 March 2013 (UTC)

Cite book comparison

Wikitext	{{cite book|isbn=978‑0812695939|title=Title}}
Live	Title. ISBN 978‑0812695939. {{cite book}}: Check |isbn= value: invalid character (help)
Sandbox	Title. ISBN 978‑0812695939. {{cite book}}: Check |isbn= value: invalid character (help)

Yep. This hack also breaks magic linking ISBN 978‑0812695939 v. ISBN 978-0812695939. The better solution would be to add markup in the template to keep this from wrapping. --— Gadget850 (Ed) ^talk 15:09, 26 March 2013 (UTC)

Please do not make any such change. Not only is it not undesirable for ISBNs to be allowed wrap, it's positively desirable, especially where cites are in 2- or 3-column format. In fact I was about to suggest that linebreaks be allowed between PMID and what follows, between doi: and what follows, and (within doi) at the slash. EEng (talk) 19:43, 26 March 2013 (UTC)

I think my reference to {{nbhyph}} is bit of a red herring. In fact it breaks if I enter the hard hyphen (&#x‍3011;) directly. John of Cromer in Philippines (talk) mytime= Wed 04:46, wikitime= 20:46, 26 March 2013 (UTC)

...(&#x‍2011;) ...! John of Cromer in Philippines (talk) mytime= Wed 04:50, wikitime= 20:50, 26 March 2013 (UTC)

non-breaking hyphen is rendered by ‑ in html. "2011" is the unicode character. 70.19.122.39 (talk) 13:51, 30 March 2013 (UTC)

hex-2011 = dec-8209 John of Cromer in Philippines (talk) mytime= Sat 22:24, wikitime= 14:24, 30 March 2013 (UTC)

{{Cite journal}}: no harvid anchor with author's name in quotes

In Metropolitan Railway there's a {{cite journal}} with the author name in quotes as it's a pseudonym, but this gives no harvid anchor

Markup	Renders as
{{cite journal|author="Fowler's Ghost"<!-- a pseudonym -->|title=Railway connections at King's Cross (part one)|magazine=[[The Railway Magazine]]|date=May 1962|ref=harv|editor-first=B.W.C|editor-last=Cooke|publisher=Tothill Press|volume=108|issue=733}}	"Fowler's Ghost" (May 1962). Cooke, B.W.C (ed.). "Railway connections at King's Cross (part one)". The Railway Magazine. 108 (733). Tothill Press. {{cite journal}}: Invalid |ref=harv (help)

whereas removing the quotes around the author works:

• Fowler's Ghost (May 1962). Cooke, B.W.C (ed.). "Railway connections at King's Cross (part one)". The Railway Magazine. 108 (733). Tothill Press. {{cite journal}}: Invalid |ref=harv (help)

Can this be fixed? Edgepedia (talk) 20:56, 26 March 2013 (UTC)

Probably. But why would you put the author's name in quotes? --— Gadget850 (Ed) ^talk 21:49, 26 March 2013 (UTC)

There is a problem with anchor encoding, but lose the quotes for this. --— Gadget850 (Ed) ^talk 02:49, 27 March 2013 (UTC)

Fixed in sandbox. Live version should be updated soon. --— Gadget850 (Ed) ^talk 12:45, 27 March 2013 (UTC)

Thank you Edgepedia (talk) 13:01, 27 March 2013 (UTC)

 Fixed --— Gadget850 (Ed) ^talk 12:24, 29 March 2013 (UTC)

Module talk:Citation/CS1

Would it make sense to merge Module talk:Citation/CS1 with this page? (I'll ask there also). Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:01, 27 March 2013 (UTC)

• Oppose. This page is for general issues about the structure of CS1 citations, whereas the talk-page Module talk:Citation/CS1 is for issues about the Lua script module, just as Template_talk:Citation/core is for issues about the markup helper template, {Citation/core}. Each talk-page focuses on different areas of concern. -Wikid77 21:33, 27 March 2013 (UTC)
  • Do you think most editors understand the distinction? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:29, 1 April 2013 (UTC)

Category:Pages using citations with accessdate and no URL

Hi, I just noted this category on a page that I edited (Journal of Reinforced Plastics and Composites). I added two references there that seem to trigger this cat. The references are online, which is why I included an accessdate. However, I cannot include a URL. The reason is that this resource is behind a pay-wall and uses dynamic session-specific URLs. This means that any URL that I would copy would not work for anyone else and, once I have closed and restarted my browser, not even for myself. I don't really see a workaround and in this case, having an accessdate but no URL seems reasonable to me. I could, of course, include a URL to the homepage of the resource, but as that is not the page where I found the information sourced by this reference, that seems less correct to me (and also runs the risk that someone will add a "fails verification" template to the reference). I'd appreciate any advice on how to handle this situation. Thanks! --Randykitty (talk) 15:34, 27 March 2013 (UTC)

Just a note for people that may be unaware. The hidden tracking category was recently added as part of the Lua migration; however, the behavior hasn't changed. It has always been the case that accessdate= is ignored if no URL is specified. Dragons flight (talk) 16:41, 27 March 2013 (UTC)

Actually, it seems to me that in the case outlined above, accessdate should perhaps not be ignored and displayed... --Randykitty (talk) 16:56, 27 March 2013 (UTC)

Yes, there is a valid question about whether accessdate should be included even if no url is given. Historically, the answer has been no, but perhaps that should change. Personally, I don't have a strong opinion either way. I was just trying to make clear that this is a discussion of a possible problem with how the accessdate parameter has been used in citations, which is different from many of the other recent reports on this page which deal with bugs specific to the Lua migration. Dragons flight (talk) 17:08, 27 March 2013 (UTC)

Per the the documentation: "Not required for web pages or linked documents that do not change; mainly of use for web pages that change frequently or have no publication date." In this instance, the access date really does not help identify the source. --— Gadget850 (Ed) ^talk 18:45, 27 March 2013 (UTC)

That's an excellent point. The URL may be dynamic, the page content isn't, so the accessdate is indeed not needed. Problem solved, I think! --Randykitty (talk) 19:01, 27 March 2013 (UTC)

No, your suspicions of the hidden-accessdate problem are correct, when no "url=" data, because the id parameters (such as "doi=") generate an internal URL which needs the accessdate to assure the time when access was allowed. The recent shake-up at Billboard.com disabled 50,000(?) weblinks. Thanks for raising the issue so quickly. -Wikid77 23:20, 27 March 2013 (UTC)

• Use parameter "postscript=" to force a date to appear: Previously, the Lua version displayed the "accessdate=" data, with no judgmental restrictions, but merely echoed all parameters as added by each user. However, with accessdate conditionally ignored, the postscript can be used to show that data:

postscript=. Retrieved 27 March 2013.

In general, ignoring parameters is typically extremely frustrating to new users who are likely to misspell parameter "url=" as perhaps capital-letter "Url=" which would be ignored, then causing the "accessdate=" value to also be hidden as well, and a new user is likely to go bonkers wondering why this "trashy" citation software does not show the URL address nor even the damned accessdate as ignoring "everything" they put in the cite. In general, there is a fine line between "smart" software and "smart-ass" software, and the automatically ignored parameters, triggered by arcane rules of citation hierarchies, will be considered by many users to be poor-quality, and non-user-friendly. Beware when the interface acts as "minefield" of traps, where one misspelled word triggers peculiar sinkholes where other parameters also disappear. In general "keep it simple" and mimic the concept of "what you see is what you get" (WYSIWYG), so if a user enters a parameter, then display that parameter without prejudiced, judgmental restrictions as to what will be deemed permissable, and instead: if they see it in the markup, they get it. However, with the speed of Lua, we will be able to process an assistance parameter (such as "help") which could warn the user, at each specific citation, about which unexpected parameters were being ignored, according to what elaborate rules for citation etiquette, or common misspellings. -Wikid77 (talk) 21:33, 27 March 2013 (UTC)

Then we should remove all restrictions and allow any and all combinations of parameters to show. We could add 'url' 'URL' 'website' 'site' 'link' 'internet' 'tubes' and for those who just can't remember 'thingy'. Then we can delete the documentation and allow editors to pick their favorite parameter names.

But really- postscript is the terminating punctuation. Yes, editors stuff all sorts of crap in it, but we aren't allowed to send them to citation reeducation camp. Unless you can come up with some sort of heuristically programmed algorithm that will sort through a free-form citation and clean it up. then we just have to live with rules. If editors can't take how the templates work then they have other options. --— Gadget850 (Ed) ^talk 22:13, 27 March 2013 (UTC)

--— Gadget850 (Ed) ^talk 22:13, 27 March 2013 (UTC)

Hey, I did not invent the way people think about computers, I just spent 5-7 years in universities, to learn what issues to beware. And "People are the way they are". Working with computer users is like a complex chess game, and to "win" their approval, there are a lot of complex issues to consider, such as trying to make complicated operations seem simple to them, without over-complicating the underlying simplicity or freedom of choices. -Wikid77 (talk) 23:20, 27 March 2013 (UTC)

Accessdate needed for doi/PMID/bibcode URL links

I found the need for the unrestricted "accessdate=" parameter, in the first article I checked ("1843 in science"), which ironically, links to an article about the "first computer program in history" (Ada Lovelace and Charles Babbage), but the cite links the webpage using a doi-parameter to generate a URL address, so the accessdate disappeared (when no "url="). Compare:

Cite journal comparison

Wikitext	{{cite journal|accessdate=2010-10-01|doi=10.1109/MAHC.2003.1253887|first1=John|first2=Jo|issue=4|journal=[[IEEE Annals of the History of Computing]]|last1=Fuegi|last2=Francis|month=October–December|pages=16–26|sandbox=true|title=Lovelace & Babbage and the creation of the 1843 'notes'|volume=25|year=2003}}
Live	Fuegi, John; Francis, Jo (2003). "Lovelace & Babbage and the creation of the 1843 'notes'". IEEE Annals of the History of Computing. 25 (4): 16–26. doi:10.1109/MAHC.2003.1253887. {{cite journal}}: |access-date= requires |url= (help); Unknown parameter |month= ignored (help); Unknown parameter |sandbox= ignored (help)
Sandbox	Fuegi, John; Francis, Jo (2003). "Lovelace & Babbage and the creation of the 1843 'notes'". IEEE Annals of the History of Computing. 25 (4): 16–26. doi:10.1109/MAHC.2003.1253887. {{cite journal}}: |access-date= requires |url= (help); Unknown parameter |month= ignored (help); Unknown parameter |sandbox= ignored (help)

Cite_quick: Template:Cite quick

Several other parameters (pmid, pmc, bibcode, OL, etc.) generate URL addresses, for which the date of access applies. See numerous URL links below:

Cite journal comparison

Wikitext	{{cite journal|accessdate=2013-03-27|bibcode=10.1109/MAHC.2003.1253887|first=Ada|issue=1|journal=Keep It Simple|last=Lovelace|month=August|ol=99A|pages=16–26|pmc=PMC999|pmid=PMID-777|sandbox=yes|title=Numerous Generated URL addresses|volume=1|year=1815}}
Live	Lovelace, Ada (1815). "Numerous Generated URL addresses". Keep It Simple. 1 (1): 16–26. Bibcode:10.1109/MAHC.2003.1253887. OL 99A. PMC 999. PMID PMID-777. {{cite journal}}: |access-date= requires |url= (help); Check |bibcode= length (help); Check |pmid= value (help); Unknown parameter |month= ignored (help); Unknown parameter |sandbox= ignored (help)CS1 maint: PMC format (link)
Sandbox	Lovelace, Ada (1815). "Numerous Generated URL addresses". Keep It Simple. 1 (1): 16–26. Bibcode:10.1109/MAHC.2003.1253887. OL 99A. PMC 999. PMID PMID-777. {{cite journal}}: |access-date= requires |url= (help); Check |bibcode= length (help); Check |pmid= value (help); Unknown parameter |month= ignored (help); Unknown parameter |sandbox= ignored (help)CS1 maint: PMC format (link)

The easiest fix is to "keep it simple" and not have any restriction to hide the "accessdate=xx" parameter. That would be another benefit of using the Lua version of the {cite_*} cites. -Wikid77 23:20, 27 March 2013 (UTC)

What does the access date mean to you? --— Gadget850 (Ed) ^talk 23:26, 27 March 2013 (UTC)

accessdate is when I or anyone else last visited the url. I hardly come across the doi etc. parameters, but aren't they as fixed as isbn, even more so, so why would they need any associated date? I'm not sure how your examples differ - I think you should subst: them in so they're fixed. Incidentally, I always specify dates as {{date|date}} because I understand that then emits date according to user's local style sheet. May also translate to user-language, I'm not sure.

John of Cromer in Philippines (talk) mytime= Thu 08:31, wikitime= 00:31, 28 March 2013 (UTC)

Apparently, people think the "accessdate=" is when they read the webpage, even if they did not know/want the URL address to include it. Perhaps that explains why many of those 40,576 people put "accessdate=" in their cites, with an empty "url=" because they are stating they accessed the page on that date. See: examples among the few 40,576 pages already in "Category:Pages using citations with accessdate and no URL". There might 90,000 people who also put the lone accessdate in {cite_web}. That is just the way people think. -Wikid77 07:29, 28 March 2013 (UTC)

I just noticed {{date}} being used in a citation. No it does not and cannot use the user date preference. With no other parameters it formats the date as DMY, thus it is rather useless here. Date formatting was dropped years ago since it relied on logged in user preferences and gave a mix of formats for casual readers. --— Gadget850 (Ed) ^talk 00:59, 28 March 2013 (UTC)

• Perhaps leave the lone accessdate until {cite_web} counted: This is an opportune time to allow the lone accessdate to remain hidden (when no "url="), until after Template:Cite_web is transitioned to Lua, to category-count all the people/pages where "accessdate=" is used without the URL parameter, among those 1.3 million {cite_web} pages. Then, based on "overwhelming demand" we can report that a consensus of "95,000" editor teams wanted to show the lone accessdate, and then perhaps sample among those thousands as to how many use "doi=" and "accessdate=" together. Remember, once {cite_web} uses Lua, then all those 1.7 million articles can be reformatted, to show accessdate, 4x times faster than any prior change to {Citation/core}. Optionally, we could upgrade to show the requested lone accessdate, but still log those pages in the Category, and later re-change to hide some of those lone accessdates. Think of this as: 3 extra chances to add new features, to the Lua cites, before incurring the overhead of 1 old update to {Citation/core}. Because once {cite_web} uses Lua, then general editing will reformat thousands of pages for "free" with the 3x faster edit-saves. -Wikid77 07:29, 28 March 2013 (UTC)

• Let's remember that the purpose of the citation is to identify the source, and each element should help in that goal. The access date is intended only for web pages that change frequently (e.g. Wikipedia main page) or have no discernible date, and has been documented that way from inception. The date someone added a citation does not help us in identifying the source. A Bibcode, doi, JSTOR or similar link does not need an access date as the content of those links by design do not change.
• Simply because editors use a field in some odd manner does not mean that we should automatically assume that there is some sort of gestalt consensus to change the use. Most casual editors will not read the documentation and will simply follow what they see in an extant page, thus bad practices are propagated. We have already seen that both page and pages are included, where the editor believes that pages is for the total number of pages in a work, which is not needed to identify the source. --— Gadget850 (Ed) ^talk 09:33, 28 March 2013 (UTC)

Yes, yes, this. 95,000 examples of improper use out 1.605 million or 1.205 million (which ever number is correct?) does not constitute consensus. Where is the discussion?

I think that Gadget850 is inadvertently helping to make my case for visible error messages (Error trapping and checks and Format but no URL) because without visible cues to the contrary, editors assume that improper use is proper and so continue with and adopt new bad practices.

—Trappist the monk (talk) 13:31, 28 March 2013 (UTC)

Error checking is good, but we need to spec it out and link to help pages for each specific issue. --— Gadget850 (Ed) ^talk 14:06, 28 March 2013 (UTC)

I've been thinking exactly that and will probably open a discussion to talk about messaging, help pages, linking from the error, etc when I return from spring break.

—Trappist the monk (talk) 23:38, 28 March 2013 (UTC)

make sure to include error-checking for when software exceeds the bounds, with overly-restrictive recommendations and the related generated errors. 70.19.122.39 (talk) 15:43, 29 March 2013 (UTC)

The Chicago Manual of Style (16th ed., p. 657–8) discusses access dates for electronic sources in general. Such a date might include electronic sources not accessible through a URL. For example, help files from software that is constantly updated. While some sort of version number might be available, the editor might not know how to access such a number, or it might be so difficult that the editor would feel the reader would be unable to access it or understand it. I admit this would be a fairly unusual occurrence. Jc3s5h (talk) 09:54, 29 March 2013 (UTC)

a version number is an edition, and should probably carry some kind of date. online sources are a subset of electronic sources, and internet/intranet sources a further subset of online sources. for such sources accessdate is the de facto edition date, if there is no other info available. 70.19.122.39 (talk) 15:43, 29 March 2013 (UTC)

• Online sources are always subject to computer-dependent revision at any time: Although stored documents might seem theoretically static, for the displayed contents, the broader reality in data processing is that all documents are subject to change when retrieved, or dynamically reformatted, from the document database(s). The problem of date-stamped materials has been so flexible that computer software, for years, has included a "build version" beyond just a computer program version number, where the contents of a "static" version might change because the underlying runtime library might have changed to alter operation of the upper-level software. For document storage, the underlying database might alter, or truncate, data in various ways, depending on unknown bugs in the current release, perhaps upgraded last week, of the database system used to display the stored document. For those reasons, an accessdate parameter helps to indicate which database or text-formatting product versions, at a point in time, were active in displaying the cited document, especially if national laws required censorship, of prior documents, for new content restrictions. -Wikid77 (talk) 19:07, 29 March 2013 (UTC)
• If an online document was subtly changed and the visible date did not change, how would you or another editor know it? --— Gadget850 (Ed) ^talk 19:23, 29 March 2013 (UTC)

well this can get complicated. if one drills down, the effects of caching may also have to be accounted for. personally i tend to think that in situations where the document (a) has not changed and (b) it is not retrieved from a cache, every access could be functionally considered a "reprint". 70.19.122.39 (talk) 00:39, 30 March 2013 (UTC)

A teaser, but it's way past my bedtime

Take a look at sharpening. There are only two citations, and they are almost identical. However the second has correctly folded the url inside itself, whereas the first one shows both the url and its title (in square brackets). What's the difference? I tried various things, such as remove format, remove pages, remove www. from the url. Nothing helped

John of Cromer in Philippines (talk) mytime= Thu 02:21, wikitime= 18:21, 27 March 2013 (UTC)

As usual, the cause of this is a newline within the linked title. Perhaps the new Lua version of the templates could find and fix this when it happens? —David Eppstein (talk) 18:29, 27 March 2013 (UTC)

This is already fixed in Lua, though we haven't yet deployed a Lua version of {{cite web}} so that page couldn't yet benefit. Dragons flight (talk) 18:45, 27 March 2013 (UTC)

 Fixed Lua deployed so this issue will be automatically fixed in the future. --— Gadget850 (Ed) ^talk 11:01, 29 March 2013 (UTC)

lua: |origyear= does not display when no author

it used to. now:

please don't ask why should anyone cite like this. probably because they have to?? let's not have a philosophical discussion again. 70.19.122.39 (talk) 01:17, 28 March 2013 (UTC)

Fixed in sandbox version. Dragons flight (talk) 01:33, 28 March 2013 (UTC)

thank you, and also thanks for the other fixes to reported problems above. 70.19.122.39 (talk) 14:21, 28 March 2013 (UTC)

 Fixed --— Gadget850 (Ed) ^talk 10:59, 29 March 2013 (UTC)

Media notes

We have three templates for citing music and DVD notes. My proposal:

• {{Cite album-notes}}: rename to 'Cite media notes' 'Cite AV media notes' and update as needed (4680 uses)
• {{Cite music release notes}}: update to 'Cite media notes' (945 uses)
• {{Cite DVD-notes}}: update to 'Cite media notes' (436 uses)

Thoughts before I take this to TfD? --  Gadget850 (Ed) ^talk 01:25, 31 March 2013 (UTC)

Sensible to merge, but is the meaning of "Cite media notes" sufficiently clear? "Cite packaging", perhaps? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:24, 1 April 2013 (UTC)

agree with the merge, but also with Andy Mabbett, in that the proposed template name may not be as precise as it should. however his "cite packaging" is too broad. the media notes referred to here are liner notes, aren't they? these pertain to the included work, not primarily to the work's medium or packaging. 70.19.122.39 (talk) 14:11, 1 April 2013 (UTC)

We recently renamed {{cite video}} to {{cite AV media}}, 'Cite AV media notes' would be more in line with that. --  Gadget850 (Ed) ^talk 14:30, 1 April 2013 (UTC)

better. 70.19.122.39 (talk) 14:34, 1 April 2013 (UTC)

Lang templates

Copied from Wikipedia talk:Help desk -- John of Reading (talk) 10:18, 1 April 2013 (UTC)

Hi. I have been using language templates in references. They were appearing really great some weeks ago. However, currently, they appear really in a bad shape. Formerly, language categories don't appear when I add the lang templates in reference titles (I think they were hidden categories). Now, they are included in the links.

For examples, see Ref # 8 in Earthquakes in 2013#References, and Ref # 4 in The Voice of the Philippines#References.--AR E N Z O Y 1 6A•t a l k• 07:13, 1 April 2013 (UTC)

Don't. We have been updating the templates. The language should have been included 'language' field without using a template.

language: The language the source is written in, if not English. Displays in parentheses with "in" before the language name. Use the full language name; do not use icons or templates.

I am sure we can fix it, but my recommendation will be to show it and put the page in an error category. The language icon templates really don't do anything but show the language in gray. Down the line, I want to use the 'language' field to properly indicate the title language. --  Gadget850 (Ed) ^talk 11:05, 1 April 2013 (UTC)

Thanks for the solution! And thanks for John of Reading for moving it here!--AR E N Z O Y 1 6A•t a l k• 12:57, 1 April 2013 (UTC)

Cite episode: producer and writer

For {{cite episode}}, not least for documentaries, we should have |producer= and |writer=. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:44, 1 April 2013 (UTC)

• 'producer' would work works, if it is followed by "producer" or "producers".
• 'writer' would work works if there are no author parameters. I can see this used if the author is a reviewer and |type=review. Then follow 'writer' with "writer" or "writers".
• Seems like this would apply to {{cite serial}} and {{cite AV media}}. --  Gadget850 (Ed) ^talk 15:09, 1 April 2013 (UTC)

Thank you. These don't seem to be documented. I'm not sure what you mean by following |producer= with "producer". Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:29, 1 April 2013 (UTC)

I refactored, as I see the confusion. You would have to identify each role; for example:

• Drucker, Sam (writer). Bloggs, Joe (producer) (January 13, 1999). "André the Giant". Biography. A&E.{{cite episode}}: CS1 maint: multiple names: authors list (link)

--  Gadget850 (Ed) ^talk 17:04, 1 April 2013 (UTC)

Although this would be more in line with Chicago:

Written by Drucker, Sam. Produced by Bloggs, Joe. (January 13, 1999). "André the Giant". Biography. A&E.{{cite episode}}: CS1 maint: multiple names: authors list (link)

--  Gadget850 (Ed) ^talk 19:47, 1 April 2013 (UTC)

Translator

'translator' has been requested multiple times. Now that we can more easily add it without a performance hit I am going to add it to the feature request at Module talk:Citation/CS1/Feature requests. --  Gadget850 (Ed) ^talk 19:50, 1 April 2013 (UTC)

Template:Cite web

Hi.

In the documentation of this template, we can read the following indication:

work: Title of website; can be wikilinked to an existing Wikipedia article or url may be used to add an external link, but not both. Displays in italics.

But when I see the Lua module, the arg work seems to be the same arg than magasine or newspaper, etc. You can see it here :

  local Periodical = args.journal or args.newspaper or args.magazine or args.work
            or args.periodical or args.encyclopedia or args.encyclopaedia

On the other hand, we can read that encyclopedia and encyclopaedia are both used for Title and Periodical. Is it normal? Is that someones knows why?

We can see it here:

    local Title = args.title or args.encyclopaedia or args.encyclopedia or args.dictionary

Thanks by advance for the answers. Automatik (talk) 20:57, 3 April 2013 (UTC)

The CS1 templates started as some 20+ separate template authored by multiple editors. Editor Gadget850 has diligently worked through all of them to make them as much the same as possible. {{Cite encyclopedia}} is the one that doesn't easily shoehorn into the others. Editor Dragons flight, author of the Lua code, has made noises about trying to fix the {{Cite encyclopedia}} mess though I suspect it's not high on the list.

Did I answer your question?

—Trappist the monk (talk) 13:07, 12 April 2013 (UTC)

Parameter="origyear"

Example	Result
{{cite web |url = http://hebrewbooks.org/38753 |title = Sepher Yezirah |origyear=Published 1877 |location=New York |work=HebrewBooks.org }}	"Sepher Yezirah". HebrewBooks.org. New York.

Why does the parameter not work? -- -- -- 00:59, 4 April 2013 (UTC)

My two penn'orth is that it should be a year, i.e. 9999, only John of Cromer in Philippines (talk) mytime= Thu 09:19, wikitime= 01:19, 4 April 2013 (UTC)

'origyear' is a child of 'year'; that is, if 'year' is not defined, then 'origyear' does not show. See the {{cite web}} documentation. The proper citation is:

Markup	Renders as
{{cite book |title=Sepher Yezirah |last=Kalisch |first=Isidor |location=New York |publisher=L. H. Frank |year=1877 |language=Hebrew |url=http://hebrewbooks.org/38753}}	Kalisch, Isidor (1877). Sepher Yezirah (in Hebrew). New York: L. H. Frank.

--  Gadget850 (Ed) ^talk 01:24, 4 April 2013 (UTC)

• Gadget850: Thanks for your quick and helpful response.
• John of Cromer: Thanks for your even quicker response. But:

1. Next time, please post your response after the end of my signature & timestamp; not before.
2. According to Template:Citation Style documentation/date, it is preferable to "supply specifics; example: |origyear=First published 1859 or |origyear=Composed 1904." -- -- -- 02:27, 4 April 2013 (UTC)

Cite web - broken display in some cases?

See reference #2 in Hennessey Venom GT:

{{cite web |url=http://www.autoblog.com/2010/03/29/hennessey-venom-gt-a-600k-mid-engine-cobra-for-the-21st-centur/ |title=Hennessey Venom GT: A $600k mid-engine Cobra for the 21st Century |accessdate=2010-03-29 |last=Lavrinc |first=Damon |date=2010-03-29 |work=[[Autoblog.com|Autoblog]] |publisher=[[Weblogs, Inc.]] }}

Gets displayed as:

Lavrinc, Damon (2010-03-29). "Hennessey Venom GT: A $600k mid-engine Cobra for the 21st Century". Autoblog. Weblogs, Inc.span class="reference-accessdate">. Retrieved 2010-03-29.

Note malformed span class. Oddly, if the publisher field ("Weblogs, Inc.") is changed to something else (e.g. "Weblogs, Inc", without the period, or "Weblogs, Inc.", without the square brackets), it works fine. GregorB (talk) 09:59, 4 April 2013 (UTC)

I suspect it is something in the new algorithm that removes extra periods. Reported. --  Gadget850 (Ed) ^talk 10:39, 4 April 2013 (UTC)

Thanks. GregorB (talk) 11:41, 4 April 2013 (UTC)

 Fixed --  Gadget850 (Ed) ^talk 19:55, 4 April 2013 (UTC)

Lavrinc, Damon (2010-03-29). "Hennessey Venom GT: A $600k mid-engine Cobra for the 21st Century". Autoblog. Weblogs, Inc. Retrieved 2010-03-29.

Press release format

In reviewing many examples of press release citations (Module talk:Citation/CS1/test/press), one of the things I noticed was that the organization issuing the press release was almost always listed a "publisher" rather than "author". For example:

{{ cite press release | publisher=HTC Corporation | title=Scartel and HTC Launch World's First Integrated GSM/WiMAX Handset | date=12 November 2008 | url=http://www.htc.com/www/press.aspx?id=76204&lang=1033 | accessdate=1 March 2011 }}

"Scartel and HTC Launch World's First Integrated GSM/WiMAX Handset" (Press release). HTC Corporation. 12 November 2008. Retrieved 1 March 2011.

and not,

{{ cite press release | author=HTC Corporation | title=Scartel and HTC Launch World's First Integrated GSM/WiMAX Handset | date=12 November 2008 | url=http://www.htc.com/www/press.aspx?id=76204&lang=1033 | accessdate=1 March 2011 }}

HTC Corporation (12 November 2008). "Scartel and HTC Launch World's First Integrated GSM/WiMAX Handset" (Press release). Retrieved 1 March 2011.

This seems a little odd to me. Obviously, there are many press releases where no specific named individuals are listed, but still the corporation is responsible for the content of the message, and personally, it seems more logical to say that "author=Big Corporation" rather than "publisher=Big Corporation". One of the reasons this might come up is that {{cite press release}} doesn't list |author= in its standard prototype (using "first" and "last" instead), so people may be unaware that "author" is an option.

If people agree that it makes sense to list the issuing corporation as the "author", it might be good to update the documentation for {{cite press release}} accordingly. Dragons flight (talk) 20:25, 5 April 2013 (UTC)

I don't really agree. I think of "author" as meaning a person. "Publisher" seems to me more appropriate for an organisation. -- Alarics (talk) 05:49, 6 April 2013 (UTC)

The flip side of this is that there are also people using "publisher" to indicate the newswire or other third party that distributed the press release. That's the more traditional meaning of publisher. It also means that some press release citations currently have no immediate indication of who wrote them unless you follow through and read the release itself. If Mom's Motors wrote a press release and Reuters distributed it, then I would say it is natural to list "Mom's Motors" as the author and "Reuters" as the publisher for essentially the same reason that if John Jacobson wrote a news article distributed by Reuters then "John Jacobson" is the author and "Reuters" is the publisher. Anyway, that's my two cents. Dragons flight (talk) 07:20, 6 April 2013 (UTC)

"Work" Parameter for Website Citation

Please explain what is meant by "Work" for Website "cite web" citations. How is it different from Author and Publisher?

I would also suggest that an explanation be given for this parameter within the Website "cite web" template, which is the only template that uses "Work" as a parameter and yet has no examples or explanations for it.Wondering55 (talk) 15:09, 6 April 2013 (UTC)

From {{cite web}}: "'work': Title of website; can be wikilinked to an existing Wikipedia article or url may be used to add an external link, but not both. Displays in italics." And many other Citation Style 1 templates use 'work' but with a slightly different meaning. --  Gadget850 (Ed) ^talk 16:26, 6 April 2013 (UTC)

Naming the illustrator or photographer in Cite book

There doesn't seem to be an obvious way to add the name of an illustrator or photographer to the Cite book template. How should it be done so that it renders "Illustrations/Photographs by Jane Doe" in the citation? Roger (talk) 11:57, 7 April 2013 (UTC)

Isn't that what the |others= field is for?Nigel Ish (talk) 12:05, 7 April 2013 (UTC)

Should the original url= be required when using archiveurl=

People here may be interested in commenting on the issue described at

Wikipedia:Village pump (policy)#Citations: Should the original url.3D be required when using archiveurl.3D. Dragons flight (talk) 18:45, 8 April 2013 (UTC)

Consistency between work and publisher parameters in citation templates

Will there ever be consistency between the work and publisher parameters in citation templates? Every Featured Article I've seen has website names in the publisher parameter of a citation template. This keeps the website from being italicized. However, {{cite web}} states that the work parameter is specifically for websites, and when using the auto-template form (not sure what to call it—appears in the edit window when you click on "Cite > Templates > Web"), the little question mark above the work parameter also states that this is for websites, which means that most people will follow this advice, since more people see the auto-template form advice than anything demanded by FA criteria. This seems to be a conflicting way of citing websites, and I'm wondering if, or what, steps have been taken to try to standardize the work/publisher parameter definitions. – Kerαunoςcopia^◁_galaxies 18:46, 8 April 2013 (UTC)

The short answer is no. 'work' is not a good parameter name in this context, and I have a feature request to add 'website' as an alias; Module talk:Citation/CS1/Feature requests. The toolbar is WP:RefToolbar. But, editors will keep doing as they have done. And many put the domain name in as the website name, which is only occasionally correct. The definitions for 'work' and 'publisher' are clearly documented.

If you want more examples, see Module talk:Citation/CS1/Rogues gallery. The first example puts the newspaper name in 'publisher' and include italics markup. --  Gadget850 (Ed) ^talk 18:55, 8 April 2013 (UTC)

I'm not sure if website as an alias would solve anything. It seems to me that users at Featured Articles are using the publisher parameter as a way to avoid the auto-italicizing. I don't believe websites should be italicized at all, that would be like italicizing "Scholastic Books". I think so long as one parameter italicizes, and the other lets you decide whether you should italicize or not, as is shown in the Rogues gallery, then editors will choose which parameter they would like to use despite any actual parameter definitions given on the template documentation. What bothers me is that users are being told to put websites in the auto-italicizing parameter, which makes no sense to me, although someone made the big decision, so I'm sure at some point there have been discussions regarding this already. But if articles that are featured on the main page don't follow this guideline, then I'm not understanding why this isn't a bigger issue on the template talk pages—hence my bringing it up here. – Kerαunoςcopia^◁_galaxies 21:46, 8 April 2013 (UTC)

Having an editor results in "In"?

[edit]Regarding {{cite book}} – Kerαunoςcopia^◁_galaxies 04:12, 11 April 2013 (UTC)
Is this new? That's very, very strange to me. What's wrong with using the "ed." abbreviation, even with an author? – Kerαunoςcopia^◁_galaxies 04:12, 11 April 2013 (UTC)

This is not new, per the template documentation:

If authors: Authors are first, followed by the included work, then "In" and the editors, 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."

--  Gadget850 (Ed) ^talk 07:39, 11 April 2013 (UTC)

Well, I saw that in the documentation, but I don't recall seeing "in" in any college papers. That just seems like a weird thing to say. Author, in another person. I've always seen "ed.", whether there's an author or not, so I'm wondering where this rule stems from. – Kerαunoςcopia^◁_galaxies 08:16, 11 April 2013 (UTC)

I can't answer for the template creators, but this is generally in line with APA Style.[2] --  Gadget850 (Ed) ^talk 15:43, 12 April 2013 (UTC)

Cite episode deprecated parameters

Editor AussieLegend has reverted this edit. With this edit, Editor AussieLegend added new text. This is the text that, prior to its deprecation, described |episodelink=.

|episodelink= was used to link |title= to a Wikipedia article. But, |episodelink=, unlike |serieslink=, has no matching |episodetitle= parameter. If the citation includes |url= and |episodelink= the resulting citation title looks like this (red link because there is no Wikipedia article titled "Episode link"):

{{cite episode |title=Episode title |episodelink=Episode link |url=http://www.example.com}}

→"Episode title". Episode link. {{cite episode}}: Missing or empty |series= (help); Unknown parameter |episodelink= ignored (|episode-link= suggested) (help)

A quick search to see how |episodelink= is used indicates that {{cite episode}} is most often used to cite other Wikipedia articles. For a spectacular example of this see Major villains in Charmed at §References.

Use of |episodelink= in this manner breaks |title= when a proper |url= is part of the citation and, when |url= is not part of the citation, serves simply as a wikilink to another article. This latter is a misuse of the citation because WP:NOTRS, particularly WP:CIRCULAR. Citations like those in Major villains in Charmed and other articles should be replaced with wikilinks in article text and proper citations to reliable sources. |serieslink= is not required because its matching parameter, |series=, can and should be wikilinked. For these reasons Editor AussieLegend's changes should be reverted and I have done so.

—Trappist the monk (talk) 14:33, 11 April 2013 (UTC)

This is not actually an accurate interpretation. The intention of |episodelink= is simply to provide a link to a Wikipedia entry on the episode. Per MOS:TV, specifically WP:TVPLOT, television episodes are acceptable primary sources. We link to the episode article because it's not possible to link to the actual episode that aired. When corrrectly used, {{cite episode}} should ideally include use of either the |minutes= or |time= parameters. For example:

• {{cite episode|title=Red-2|episodelink=#ep91|series=NCIS: Los Angeles|serieslink=NCIS: Los Angeles|network=[[CBS]]|date=March 26, 2013|season=4|number=19|minutes=02:21}} produces
  "Red-2". NCIS: Los Angeles. Season 4. Episode 19. March 26, 2013. 02:21 minutes in. CBS. {{cite episode}}: Unknown parameter |episodelink= ignored (|episode-link= suggested) (help); Unknown parameter |serieslink= ignored (|series-link= suggested) (help)

The documentation as I modified it simply reflects the way that the template works. It does not change the way that the template works; |episodelink= is still a functional parameter after Trappist the monk's removal of the documentation, which was modified around this time last year to change the description of |title= to "Title of source. Can be wikilinked to an existing Wikipedia article or url may be used to add an external link, but not both. Displays in quotes." (This was how the parameter was deprecated. There doesn't appear to have been any discussion about deprecating either |episodelink= or |serieslink=.) Despite the "but not both" claim, it is still possible both link to an article in the title and use a url. Doing so results in:

• {{cite episode|title=[[NCIS: Los Angeles (season 4)#ep91|Red-2]]|url=http://www.example.com}} producing:
  "[[NCIS: Los Angeles (season 4)#ep91|Red-2]]". {{cite episode}}: Missing or empty |series= (help); URL–wikilink conflict (help)

or, to use Trappist the monk's example:

• "[[Episode link|Episode title]]". {{cite episode}}: Missing or empty |series= (help); URL–wikilink conflict (help)

This results in the same peculiar output as Trappist the monk's example, where both the url and the wikilink are shown, with the url preceding the wikilink. Argue as you may that it won't, history shows that the likelihood such a citation will be added is high - I'm always fixing incorrect uses of this template. The solution would seem to be simply to modify {{cite episode}} so that |url= overrides |episodelink=; hiding the problem with documentation hacks is not the way to do it. I suspect that the reason this was never done previously is that, being a broadcast medium, there rarely is a url to a particular episode. Other than a recap on an official site, any such url is generally a linkvio and most urls don't actually seem to support the information that is being cited. There is no explanation as to why serieslink has been deprecated. It too is still a fully functional parameter. --AussieLegend (✉) 16:15, 11 April 2013 (UTC)

Stepping back from the issue of what it does output, which is pretty clearly malformed, do you have an opinion on what ought to happen if both episodelink and url are specified? If you do something like this with the new Lua citations you get:

{{cite web|title=Episode title |titlelink=Episode link |url=http://www.example.com}}

→"Episode title". {{cite web}}: Unknown parameter |titlelink= ignored (|title-link= suggested) (help)

Which is arguably better than the present format, but probably still bad because there is no description on the URL. Dragons flight (talk) 17:07, 11 April 2013 (UTC)

As I said, "The solution would seem to be simply to modify {{cite episode}} so that |url= overrides |episodelink=". In the event that a url is available, you should see "Episode title". {{cite episode}}: Missing or empty |series= (help), which is easier to do if we continue to use |episodelink=. --AussieLegend (✉) 17:24, 11 April 2013 (UTC)

Having said that, I've discovered a now archived proposal by 117Avenue that may be an alternative. I haven't checked it out thoroughly, but you can find it here. --AussieLegend (✉) 17:32, 11 April 2013 (UTC)

(edit conflict)|serieslink= was and should be deprecated because:

|series=NCIS: Los Angeles|serieslink=NCIS: Los Angeles

is the same as:

|series=[[NCIS: Los Angeles]]

This works because |series= is not a synonym for another parameter and technically identical to how you implemented |network=.

How about this? Wikilink to the Wikipedia episode article through the |number= parameter. Doing that shows how |episodelink= isn't necessary, reduces duplication, and leaves |title= free for use with |url=. Here is a tweaked version of your example:

{{cite episode|title=Title for link to external NCIS: Los Angeles (season 4) site|url=http://www.example.com|series=[[NCIS: Los Angeles]]|network=[[CBS]]|date=March 26, 2013|season=4|number=[[NCIS: Los Angeles (season 4)#ep91|19]]|minutes=02:21}}

→"Title for link to external NCIS: Los Angeles (season 4) site". NCIS: Los Angeles. Season 4. Episode 19. March 26, 2013. 02:21 minutes in. CBS.

Presumably |url= identifies a WP:RS.

Per MOS:TV, specifically WP:TVPLOT, television episodes are acceptable primary sources. Well, yeah ... Except that it states in the same paragraph: "Since the episode is the primary source and the infobox provides details about it, citing the episode explicitly in the plot summary's section is not necessary." I think that this supports my position that the citations in articles like Major villains in Charmed should be reduced to wikilinks in the text rather than malformed "citations".

Here is a {{cite web}} version of the tweaked {{cite episode}} citation above:

{{cite web|title=Title for link to external [[NCIS: Los Angeles]] (season 4) site|url=http://www.example.com|series=[[NCIS: Los Angeles]]|network=[[CBS]]|date=March 26, 2013|season=4|number=[[NCIS: Los Angeles (season 4)#ep91|19]]|minutes=02:21}}

→"Title for link to external [[NCIS: Los Angeles]] (season 4) site". NCIS: Los Angeles. CBS. March 26, 2013. 02:21 minutes in. {{cite web}}: URL–wikilink conflict (help); Unknown parameter |nocat= ignored (|no-tracking= suggested) (help); Unknown parameter |season= ignored (|date= suggested) (help)

The reason I show you this is because this is how Module:Citation/CS1 treats the case where |title= has both |url= and a competing wikilink. This is the future. CS1 will eventually handle all of the 23 or so separate Citation Style 1 templates.

—Trappist the monk (talk) 17:38, 11 April 2013 (UTC)

|serieslink= is often not the same as |series=. NCIS won't get you the TV series, Hell's Kitchen won't get you to the UK TV series or the U.S. TV series. The statement "|serieslink= is not required because its matching parameter, |series=, can and should be wikilinked" is not entirely correct either. The new documentation only says, "The name of the series the episode belongs to; may be wikilinked" but the old documentation gave more detail: "If the citation is being used in the article about the series itself, this call is not necessary and will in fact create improper formatting." This is more correct - when used in the main article, the series title will be improperly bolded in the citation:

• "Red-2". NCIS: Los Angeles. Season 4. Episode 19. March 26, 2013. 02:21 minutes in. CBS. {{cite episode}}: Unknown parameter |episodelink= ignored (|episode-link= suggested) (help)

i.e. The series name should not be linked in some circumstances. I know from experience that when re-using a citation from a series or episode article, or episode list, which happens very often, it's a lot more convenient to strip "|serieslink=Series name (TV series)" from a citation than it is to delink [[Series name (TV series)|Series name]]. OK, it's not extremely complicated but I don't see the point in making anything more difficult than it needs to be and this has been working fine for years. |network= rarely requires disambiguation, just linking, which is why there is no |networklink=.

""Since the episode is the primary source and the infobox provides details about it, citing the episode explicitly in the plot summary's section is not necessary." I think that this supports my position that the citations in articles like Major villains in Charmed should be reduced to wikilinks in the text rather than malformed "citations"." - No, I'm afraid it doesn't. The section that you've quoted only applies to episode articles that have an infobox for the specific episode. Episode lists have a different infobox and {{cite episode}} isn't normally used in the plot section of episode lists or articles, although sometimes it is necessary. More often than not it's used in TV series main articles, character articles and even some completely unrelated articles that need to cite a television episode. It provides for the inclusion of times using |minutes= or |time=, which a wikilink doesn't. These parameters point to a specific time in the episode that supports the claim. For example, prior to airing, the NCIS: Los Angeles episode "Red-2" was listed in reliable sources as "Red: Part Two".[3] However, when it aired the title shown on-screen was "Red-2". The citation that I've been using as an example includes the time in the episode ("02:21 minutes in") where this is shown. Unlike a newspaper or website, where a reader can usually do a word search to verify a claim, this is not possible in a TV episode. A time code is needed in order to avoid having to watch a large chunk of most episodes (not all information is in the opening credits and fast-forwarding often misses the information, especially when it is spoken). --AussieLegend (✉) 17:58, 11 April 2013 (UTC)

Some sort of (edit conflict) occurred here that resulted in part of Editor AussieLegend's above post getting deleted. Not sure how that happened; there was no warning at the time I saved my post below. I think that Editor AussieLegend's has been restored,

—Trappist the monk (talk) 19:20, 11 April 2013 (UTC)

It isn't clear to me why there are any |whateverlink= parameters except for |authorlink= and |editorlink= - these because author and editor names are often broken up into given- and surnames. Your argument, as I understand it, is that creation and maintenance of disambiguated wikilinks is an onerous task. I'm not persuaded.

I'm not understanding the need to cling to {{cite episode}} and |episodelink= when all that it is accomplishing is a multiple-step link from the article text that the reader is reading: (click) to §References where the reader has to figure out which link in the citation to follow; (click) to the episode article. Tell me why using {{cite episode}} and |episodelink= is better for this than a simple wikilink in the article text where the reader is reading; (click) to the episode article.

Documentation can pretty much always be made more clear just as what we write here in this discussion can be made more clear. Perhaps I should have written "|serieslink= is not required because its matching parameter, |series=, can and should be wikilinked instead."

—Trappist the monk (talk) 18:49, 11 April 2013 (UTC)

(taps himself on the shoulder and reminds himself that the topic of the discussion is deprecated parameters)

This cite does what I think you want yet doesn't use the deprecated parameters |episodelink= and |serieslink=:

{{cite episode|title=Red-2|series=[[NCIS: Los Angeles]]|network=[[CBS]]|date=March 26, 2013|season=4|number=[[NCIS: Los Angeles (season 4)#ep91|19]]|minutes=02:21}}

→"Red-2". NCIS: Los Angeles. Season 4. Episode 19. March 26, 2013. 02:21 minutes in. CBS.

—Trappist the monk (talk) 19:36, 11 April 2013 (UTC)

"Your argument, as I understand it, is that creation and maintenance of disambiguated wikilinks is an onerous task. I'm not persuaded." - May I ask how many television articles you've edited and had to reference using {{cite episode}}? If the answer is what I think, then I can understand why you can't see. As an editor who has edited hundreds of TV articles I can say that it is something that saves considerable time and, as I've said, it's something that has worked well for years, so I don't see what is to be gained by deprecating useful parameters.

"Tell me why using {{cite episode}} and |episodelink= is better for this than a simple wikilink in the article text where the reader is reading; (click) to the episode article." - The episode link adds information. A link to the episode entry isn't really needed in the citation at all, if editors use the |time= or |minutes= parameters, but they are parameters that are often omitted - editors can often cite something from an episode but don't have the time from the episode that something occurred. Of course we could say that most citation parameters aren't needed. All we really need is <ref>http://www.example.com</ref> but we encourage editors to add as much information as possible to citations. Where a detailed episode article exists, a link to the episode article can often provide the information needed, without forcing the reader to have to search for a copy of the episode, especially if it's something that just deals with plot information. The full citation information is still present, but the link often provides a shortcut. It's also not always practical to provide a wikilink to an episode in the prose. Sometimes the episode title just doesn't "fit" - "The warden was revealed to be a man.^[10]" is preferable to "As stated in "Caged Fae", The warden was revealed to be a man.") When an episode is referenced multiple times, as often happens, a wikilink in each place generally violates WP:REPEATLINK. An episodelink in the citation avoids that.

"Perhaps I should have written "|serieslink= is not required because its matching parameter, |series=, can and should be wikilinked instead."" - No, because that's wrong too. There are times, as explained above, when | should NOT be wikilinked.

"This cite does what I think you want yet doesn't use the deprecated parameters" - It does, but I've explained the problems with that above, and you haven't convinced me that those parameters should be deprecated, remembering that never was any discussion that lead to them being deprecated. There was just one arbitrary edit that decided they should be deprecated. It was a bold edit, but it has been opposed. --AussieLegend (✉) 13:02, 12 April 2013 (UTC)

I updated the template to {{citation/core}} over a year ago and the documentation a bit later.discussion My reasoning for deprecating, but not removing these two parameters was to encourage editors to use the wikilink method of the other Citation Style 1 templates but retain the parameters for backward compatibility.

As we move forward in updating CS1 templates to use Lua, we have added a lot more error checking. For example, with {{cite book}}, if you specify 'url' and use a wikilink in 'title' then there is an error message. These are not yet visible as we want the help system to be in place before editors start seeing a bunch of new errors. This particular check would not fix the issue with 'serieslink' and a wikilink in 'title', but it would give an error and place the page in a category.

I have added a feature request to check to see if a wikilink is to the current page and simply kill the link. --  Gadget850 (Ed) ^talk 16:20, 12 April 2013 (UTC)

--  Gadget850 (Ed) ^talk 16:20, 12 April 2013 (UTC)

What you guess is probably close to right, but the number of whatever-type of article edits I've made has no bearing on whether or not |episodelink= and |serieslink= should be deprecated. Have you considered AWB for these repetitive edits?

Have you considered using {{rp}} to reference a specific event time?

However, when it aired the title shown on-screen was "Red-2".

I included a linked title which may not be the right destination but you get the point.

In your Caged Fae example, does the reader not already know that the subject is the Lost Girl Caged Fae episode? If not, then perhaps the article text needs editing to make that clear. While it might be helpful to the reader to know that the discussion is about the Caged Fae episode, it may be inconvenient or inappropriate to include that information at that particular place in the article text. So then perhaps this:

The warden was revealed to be a man."

or something similar. The sentence might need rewording to more fully accommodate the wikilink. I don't think that this violates the tenants of WP:REPEATLINK.

Since I haven't convinced you and you haven't convinced me, here we stand unresolved. So, I've been wondering if an alternate solution exists that answers the need to make reference to an episode article, event time, etc that doesn't use {{cite episode}} at all so there is no need for |episodelink= and |serieslink=. What if we were to create a referencing template {{ref episode}}? It might provide <ref group=episode></ref> tags or similar so that references to episode articles would be automatically grouped together under a separate {{reflist|group=episode}}. Parameters might include |title=, |number=, |minutes=, etc. No external link conflicts; |title= and |number= are wikilinkable. Not sure how to do <ref group=episode name="??" /> but I'm sure it can be figured out – {{sfn}} does it which might be forked to create {{ref episode}}. Here, in this thread though, is not the place to discuss the details of this {{ref episode}} idea.

User:Trappist the monk|Trappist the monk]] (talk) 16:03, 12 April 2013 (UTC)

If you haven't been through the experience, then it's hard to understand what it means to have the convenience. That's not a criticism of you, it just makes it hard to convey the information. For example, most people wouldn't know the advantage the pedals in a left hand drive Lamborghini Gallardo has over a right hand drive model (just trying to pick an "out there" example"). No, I haven't considered using {{rp}} because it is "is for appending page numbers in Harvard referencing style (or AMA style", not for adding time codes. {{Cite episode}} is a template specifically for citing episodes including times within episodes. We really should use the right tool for the right job, and the right tool here is {{Cite episode}}.

"In your Caged Fae example, does the reader not already know that the subject is the Lost Girl Caged Fae episode?" - Yes, of course they should. I was attempting to show you how episode titles don't necessarily fit into the prose - don't overthink it.

""The warden was revealed to be a man."" - That's pretty ambiguous. Wikilinks should be obvious in their purpose. A reader is more likely to ignore such a link, thinking that it's a link to an article related to "revealed".

"So, I've been wondering if an alternate solution exists that answers the need to make reference to an episode article, event time, etc that doesn't use {{cite episode}} at all so there is no need for |episodelink= and |serieslink=." - Again, don't overthink it. The reality is that, there was never any consensus to deprecate episodelink and serieslink and they remain fully functional parameters. They've only disappeared from the documentation, again without any consensus to remove them. The template is transcluded 7,226 times and many uses include episodelink and serieslink. There is no reason why we shouldn't continue to use them. What we DO need to do is fix the issue caused when the episode title is linked (in whatever form it is linked) and |url= is present. I don't think anyone disagrees with the argument that url should override wikilinking. --AussieLegend (✉) 16:46, 12 April 2013 (UTC)

The parameters are still supported, so use them as desired. The issues can be resolved per my post just above, unless there is an objection to the Lua update. --  Gadget850 (Ed) ^talk 17:31, 12 April 2013 (UTC)

Need help

Hi,

Is there someone to take care of my question : #Template:Cite web? Automatik (talk) 11:36, 12 April 2013 (UTC)

Done.

—Trappist the monk (talk) 13:08, 12 April 2013 (UTC)