Template talk:Citation/Archive 4: Difference between revisions

Content deleted Content added

Inline

Revision as of 01:42, 31 August 2016

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

→

Identifier overhaul

{{editprotected}} Not sure what exactly should be made, since I'm not intimately familiar with the template. Headbomb {^ταλκ_{κοντριβς} – WP Physics} 22:52, 1 March 2010 (UTC)

In that case the {{editprotected}} is premature. Please place the code in the sandbox before placing the request. Thanks — Martin (MSGJ · talk) 08:46, 2 March 2010 (UTC)

~~Can anyone implement this in the sandbox?~~ Nevermind, it's actually much easier than I thought. Headbomb {^ταλκ_{κοντριβς} – WP Physics} 17:37, 2 March 2010 (UTC)

Done and tested. However there might not be consensus for this (see CODGEN's concerns on transclusions), and I don't pretend to be neutral on the issue, since I have a COI for a rapid fix. Headbomb {^ταλκ_{κοντριβς} – WP Physics} 19:39, 2 March 2010 (UTC)

Hi. I had sandboxed some changes to Citation/core prior to your changes, but I left them to be looked at while I was away on a trip and your changes seem to have overtaken mine before they were published. See the last few entries in #Coauthors above for discussion about that. I see from your remarks above that there is uncertainty re consensus for your changes. Unless asked to hold off, I'll probably redo my changes in a day or two using the current Citation/core version as a starting point. If that is a problem, please let me know. Wtmitchell (talk) (earlier Boracay Bill) 04:53, 13 March 2010 (UTC)

I've been thinking it would be much easier to maintain all the various identifier if we created a sort of {{citation/identifier}} template, in which the various identifiers could be streamlined (or amalgamated). For example, the doi, arxiv, and PMC identifiers could be called as

Input

Output

{{citation/identifier
 |identifier=doi
 |input1={{{doi|10.foobar}}}
}}

doi:10.foobar

{{citation/identifier
 |identifier=arxiv
 |input1={{{eprint|}}}
 |input2={{{version|}}}
 |input3={{{class|}}}
}}

arXiv:{{{eprint}}}

{{citation/identifier
 |identifier=pmc
 |input1={{{pmc|}}}
}}

PMC {{{pmc}}}

or something. Adding and removing identifiers would simply become a matter of adding

{{citation/identifier
 |identifier=foo
 |input1={{{foo|}}}
}}

in the identifer section of {{citation}} (or {{citation/core}}, whichever). And maintenance on the identifier themselves would be very easy. Opinions? Headbomb {^ταλκ_{κοντριβς} – WP Physics} 04:19, 28 February 2010 (UTC)

As long as it preserves the individual formatting styles of the different kinds of identifiers, I don't see why not. —David Eppstein (talk) 04:24, 28 February 2010 (UTC)

Alright, I've coded arxiv, doi, ASIN, bibcode, ISBN, ISSN, JFM, JSTOR, MR, OCLC, PMC, PMID, and ZBL. Pretty sure they're all identical to what's in right now, but it wouldn't hurt to check. This would also take care of DOI tweak for print (and possibly other IDs).Headbomb {^ταλκ_{κοντριβς} – WP Physics} 05:14, 28 February 2010 (UTC)

This template could also be used to streamline the various identifiers template such as {{issn}} and {{jstor}}. Headbomb {^ταλκ_{κοντριβς} – WP Physics} 06:07, 28 February 2010 (UTC)

I've been having thoughts about something else which would also lead to multiple template transclusions in {{citation/core}}, and I've had vague concerns about performance impact (see Wikipedia:Transclusion costs and benefits#Double transclusion). I know that there was a big issue some years back related to this. Is this known to now not be an issue? If not, perhaps this should be explored before too much implementation work is done. Wtmitchell (talk) (earlier Boracay Bill) 06:00, 28 February 2010 (UTC)

Changes are rolled out slowly so it doesn't tax the servers beyond what they can handle, so this is pretty much a de facto non-issue as far as I'm aware. Also, the point of these subtemplates is that they affect lots of pages, so one only needs to change one page to affect several thousands, if not millions of them. This ensures uniformity throughout the project, which is the point of templates. Headbomb {^ταλκ_{κοντριβς} – WP Physics} 06:07, 28 February 2010 (UTC)

And my point was that an article which uses {{citation}} or {{cite xxx}} citation/core has (1 + 1)*(several hundred) transclusions to get citation/core in. If citation/core then does five internal transclusions, that changes to (1 + (5*(several hundred))) transclusions. If template transclusion has a nontrivial performance impact, this might be significant. As I recall, it was significant enough at some point in the past for a limit to be placed on the total number of transclusions processed in rendering a page. I don't know whether or not such a limit is still in place.

It now occurs to me, though, that a way to deal with this would be to add an inline expansion step into the process of publishing sandboxed changes — as a final step in publishing citation/core/sandbox changes, all subtemplates internally transcluded there could be subst'd, so there would be no transclusions done internally by the published version of citation/core. Wtmitchell (talk) (earlier Boracay Bill) 06:49, 28 February 2010 (UTC)

Not convinced it's a good idea. But as long as the identified thing is streamlined, and the print errors fixed, I don't care about the details of implementation, so I'll defer to those who'll implement it. Headbomb {^ταλκ_{κοντριβς} – WP Physics} 08:16, 28 February 2010 (UTC)

Alright, I've made the changes to Template:Citation/core/sandbox and tested everything in print and online. I haven't seen any change in behaviour online, and the print is now much more sane for the identifiers. See User:Headbomb/Sandbox and its PDF.

However, it would still be a good idea for someone to double check my changes. See diffs. Also note that some author-related work has been done on the sanbox recently, so that shouldn't be uploaded along with my changes if they are deemed acceptable. Headbomb {^ταλκ_{κοντριβς} – WP Physics} 18:51, 2 March 2010 (UTC)

I'm skeptical of adding transclusions unless there is a significant improvement in simplicity or some other tangible advantage. The original division between Citation and Citation/core, for example, was necessary because it allowed synonym substitution and other basic parameter processing to take place once in Citation, instead of multiple times in Citation/core. This actually improves performance. I don't see any similar advantage to breaking out the identifier processing into a separate template. I do think we ought to be very sensitive to performance, because some articles contain scores of references to Citation-based templates, and there can be a noticeable delay in loading these articles, though it's not as bad as it was a few years ago. COGDEN 19:00, 2 March 2010 (UTC)

Well, these can be subst: or whatever, but the fact remains that currently things in print are broken (or at very least incredibly ugly [the problem is much deeper than simply ^[1] boxes next to identifiers]). Optimization can always be performed later on if it's found to be a problem. Headbomb {^ταλκ_{κοντριβς} – WP Physics} 19:08, 2 March 2010 (UTC)

I made the edit protect request, trusting that MSGJ (or others) would know if this is actually a solution which introduces a performance hit not worth the problem that it would solve. If the performance hit is not worth it, back to the drawing board, otherwise, let's implement it. Headbomb {^ταλκ_{κοντριβς} – WP Physics} 19:43, 2 March 2010 (UTC)

Note that a discussion is currently going on about this on the WP (Wikipedia:Village pump (proposals)#Tweak of identifier in print). Headbomb {^ταλκ_{κοντριβς} – WP Physics} 22:59, 3 March 2010 (UTC)

Problem with editor parameter

It seems that when material has an editor, but no author, the name of the editor doesn't show up (see the second example in Template:Citation#Examples_3). That needs to be fixed. Kaiser matias (talk) 05:45, 20 March 2010 (UTC)

A similar problem has been reported at Template talk:Cite book. Cavila (talk) 12:46, 20 March 2010 (UTC)

I think we'd better centralise the discussion of these related problems, I suggest the talk page of the template which is at fault, ie at Template talk:Citation/core#Last change breaks Editor fields. --Redrose64 (talk) 14:09, 20 March 2010 (UTC)

Date duplication since change

I noticed a change in quite a number of references where no author is specified, the date is now displayed at the beginning of the citation in parentheses and later in the citation after the page number. This is broken behavior which I believe traces back to a change made to {{Citation/core}} in the last 12 hours or so. As it's protected and I can't be absolutely sure, I'm hoping somebody more intimately familiar with template syntax can determine the actual cause of the problem and eliminate it. - Dravecky (talk) 08:43, 20 March 2010 (UTC)

I think we'd better centralise the discussion of these related problems, I suggest the talk page of the template which is at fault, ie at Template talk:Citation/core#Last change breaks Editor fields. --Redrose64 (talk) 14:08, 20 March 2010 (UTC)

Parameter work= is not documented, and has strange behavior

The work= parameter is not documented, but I assume it has the same meaning as in {{cite web}}, to name a web site etc.

{{citation|title=The state of the web|work=news.com}}

yields

"The state of the web", news.com

as expected, but

{{citation|title=The state of the web|newspaper=|work=news.com}}

gives

"The state of the web", news.com.

I don't think an empty newspaper= parameter should make that much of a difference; I often copy and paste the template with all its parameters and leave the ones that don't apply empty. AxelBoldt (talk) 17:31, 28 March 2010 (UTC)

Depending upon how a template is written, leaving a parameter blank will not necessarily give the same results as omitting the parameter entirely. Generally speaking, it's best to remove the ones for which there is no relevant value; however if you believe that a suitable value should be put there, but cannot as yet find it, leave it as an empty parameter as a reminder.

Checking the source for {{citation}}, I find that |journal=, |periodical=, |newspaper=, |magazine= and |work= are mutually exclusive: this is because they are passed through to {{citation/core}} via the latter's |Periodical= parameter. If more than one of these is present (whether empty or not), the earliest one in that list will be used, and all of the others which are after it in the list will be ignored. Thus, by specifying both |newspaper= and |work=, |newspaper= is used and |work= is ignored, because |newspaper= has precedence. --Redrose64 (talk) 18:21, 28 March 2010 (UTC)

Would it be possible to only pass non-empty parameters through to citation/core's Periodical parameter? Right now I can't see a use-case where the current behavior would be preferable, but maybe I'm missing something. AxelBoldt (talk) 04:26, 29 March 2010 (UTC)

titleurl

For anyone who is interested and is unaware of this, I want to mention the discussions at Template talk:Cite book#Chapter and url and at Template talk:Citation/core#Titles, IncludedWorks, and URLs. If the changes discussed there are implemented, changes similar to the cite book changes would probably also be applied here in the interests of consistency. Wtmitchell (talk) (earlier Boracay Bill) 04:42, 22 April 2010 (UTC)

author-link and weird display?

At Raley Field the first reference looks like: "[|Knigh, Graham]" rather than just "Knigh, Graham" - why is that?

{{Citation | last = Knigh | first = Graham | author-link = http://www.baseballpilgrimages.com/ | last2 = Smight | first2 = Tim | author2-link = | title = 10 great places for a baseball pilgrimage | date = 2010/04/01 | year = 2010 | url = http://www.usatoday.com/travel/destinations/10great/2010-04-01-minor-league-ballparks_N.htm | accessdate = 2010/05/03}}

as compared to

{{Citation | last = Knigh | first = Graham | author-link = | last2 = Smight | first2 = Tim | author2-link = | title = 10 great places for a baseball pilgrimage | date = 2010/04/01 | year = 2010 | url = http://www.usatoday.com/travel/destinations/10great/2010-04-01-minor-league-ballparks_N.htm | accessdate = 2010/05/03}}

Adding the "author-link" URL generates the extra characters it seems. j-beda (talk) 17:41, 3 May 2010 (UTC)

That is because it is only for links to Wikipedia articles on the author, not external links. For example, setting |last=Jeal |first=Tim |author-link=Tim Jeal would create Jeal, Tim. ---— Gadget850 (Ed) ^talk 20:20, 3 May 2010 (UTC)

Format broken by ] character in DOI

It appears that the use of [ and ] characters can be problematic in a DOI - see this doi: 10.1642/0004-8038(2006)123[475:WMAGNB]2.0.CO;2 as used in Baya Weaver (reference number 32). The link seems fine but the appearance is somewhat odd. Shyamal (talk) 07:45, 21 March 2010 (UTC)

This should be fixed now. Headbomb {talk / contribs / physics / books} 23:14, 4 May 2010 (UTC)

Fix

To fix problems such as these, replace all {{{input1|}}}] with {{#tag:nowiki|{{{input1|}}}}}], and all {{{input1|}}}{{{input2|}}}] with {{#tag:nowiki|{{{input1|}}}{{{input2|}}}}}]. AKA replace

current

{{#switch:{{{identifier}}}
 |arxiv={{hide in print
         |[[arXiv]]:[http://arxiv.org/abs/{{{input1|}}}{{{input2|}}} {{{input1|}}}{{{input2|}}}]<!--
          -->{{#if:{{{input3|}}}
              |&nbsp;[{{{input3|}}}]
              |
             }}
        }}<!--
     -->{{only in print
         |arXiv:{{{input1|}}}{{{input2|}}}<!--
          -->{{#if:{{{input3|}}}
              |&nbsp;[{{{input3|}}}]
              |
             }}
        }}
 |doi={{hide in print
       |{{#if:{{{input2|}}}
         |[[Digital object identifier|doi]]:{{{input1|}}} (inactive {{{input2|}}}) {{#ifeq: {{NAMESPACE}} | {{ns:0}}
                   |[[Category:Pages with DOIs broken since {{#time: Y | {{{input2|}}} }}]]
                   |
                  }}
         |[[Digital object identifier|doi]]:[http://dx.doi.org/{{urlencode:{{{input1|}}}}} {{{input1|}}}]
        }}
      }}<!--
   -->{{only in print
       |{{#if:{{{input2|}}}
         |doi:{{{input1|}}} (inactive {{{input2|}}})
         |doi:{{{input1|}}}
        }}
      }}
 |asin={{hide in print
        |[[Amazon Standard Identification Number|ASIN]]&nbsp;[http://www.amazon.com/dp/{{{input1|}}} {{{input1|}}}]
       }}<!--
    -->{{only in print
        |ASIN&nbsp;{{{input1|}}}
       }}
 |bibcode={{hide in print
           |[[Bibcode]]:&nbsp;[http://adsabs.harvard.edu/abs/{{{input1|}}} {{{input1|}}}]
          }}<!--
       -->{{only in print
           |Bibcode:&nbsp;{{{input1|}}}
          }}
 |jfm={{hide in print
       |[[Jahrbuch über die Fortschritte der Mathematik|JFM]]&nbsp;[http://www.zentralblatt-math.org/zmath/en/search/?q{{=}}an:{{{input1|}}}&format{{=}}complete {{{input1|}}}]
      }}<!--
   -->{{only in print
       |JFM&nbsp;{{{input1|}}}
      }}
 |jstor={{hide in print
         |[[JSTOR]]&nbsp;[http://www.jstor.org/stable/{{{input1|}}} {{{input1|}}}]
        }}<!--
     -->{{only in print
         |JSTOR&nbsp;{{{input1|}}}
         }}
 |isbn={{hide in print
        |[[International Standard Book Number|ISBN]]&nbsp;[[Special:BookSources/{{{input1|}}}|{{{input1|}}}]]
       }}<!--
    -->{{only in print
        |ISBN&nbsp;{{{input1|}}}
       }}
 |issn={{hide in print
        |[[International Standard Serial Number|ISSN]]&nbsp;[http://www.worldcat.org/issn/{{{input1|}}} {{{input1|}}}]
       }}<!--
    -->{{only in print
        |ISSN&nbsp;{{{input1|}}}
       }}
 |mr={{hide in print
      |[[Mathematical Reviews|MR]][http://www.ams.org/mathscinet-getitem?mr{{=}}{{{input1|}}} {{{input1|}}}]
     }}<!--
  -->{{only in print
      |MR{{{input1|}}}
     }}
 |oclc={{hide in print
        |[[Online Computer Library Center|OCLC]]&nbsp;[http://www.worldcat.org/oclc/{{{input1|}}} {{{input1|}}}]
       }}<!--
    -->{{only in print
        |OCLC&nbsp;{{{input1|}}}
       }}
 |pmc={{hide in print
       |[[PubMed Central|PMC]]&nbsp;[http://www.pubmedcentral.gov/articlerender.fcgi?tool{{=}}pmcentrez&artid{{=}}{{{input1|}}} {{{input1|}}}]
      }}<!--
   -->{{only in print
       |PMC&nbsp;{{{input1|}}}
      }}
 |pmid={{hide in print
        |[[PubMed Identifier|PMID]]&nbsp;[http://www.ncbi.nlm.nih.gov/pubmed/{{{input1|}}} {{{input1|}}}]
       }}<!--
    -->{{only in print
        |PMID&nbsp;{{{input1|}}}
       }}
 |zbl={{hide in print
        |[[Zentralblatt MATH|ZBL]]&nbsp;[http://www.zentralblatt-math.org/zmath/en/search/?q{{=}}an:{{{input1|}}}&format{{=}}complete {{{input1|}}}]
       }}<!--
    -->{{only in print
        |ZBL&nbsp;{{{input1|}}}
       }}
}}

with

new

{{#switch:{{{identifier}}}
 |arxiv={{hide in print
         |[[arXiv]]:[http://arxiv.org/abs/{{{input1|}}}{{{input2|}}} {{#tag:nowiki|{{{input1|}}}{{{input2|}}}}}]<!--
          -->{{#if:{{{input3|}}}
              |&nbsp;[{{{input3|}}}]
              |
             }}
        }}<!--
     -->{{only in print
         |arXiv:{{{input1|}}}{{{input2|}}}<!--
          -->{{#if:{{{input3|}}}
              |&nbsp;[{{{input3|}}}]
              |
             }}
        }}
 |doi={{hide in print
       |{{#if:{{{input2|}}}
         |[[Digital object identifier|doi]]:{{{input1|}}} (inactive {{{input2|}}}) {{#ifeq: {{NAMESPACE}} | {{ns:0}}
                   |[[Category:Pages with DOIs broken since {{#time: Y | {{{input2|}}} }}]]
                   |
                  }}
         |[[Digital object identifier|doi]]:[http://dx.doi.org/{{urlencode:{{{input1|}}}}} {{#tag:nowiki|{{{input1|}}}}}]
        }}
      }}<!--
   -->{{only in print
       |{{#if:{{{input2|}}}
         |doi:{{{input1|}}} (inactive {{{input2|}}})
         |doi:{{{input1|}}}
        }}
      }}
 |asin={{hide in print
        |[[Amazon Standard Identification Number|ASIN]]&nbsp;[http://www.amazon.com/dp/{{{input1|}}} {{#tag:nowiki|{{{input1|}}}}}]
       }}<!--
    -->{{only in print
        |ASIN&nbsp;{{{input1|}}}
       }}
 |bibcode={{hide in print
           |[[Bibcode]]:&nbsp;[http://adsabs.harvard.edu/abs/{{{input1|}}} {{#tag:nowiki|{{{input1|}}}}}]
          }}<!--
       -->{{only in print
           |Bibcode:&nbsp;{{{input1|}}}
          }}
 |jfm={{hide in print
       |[[Jahrbuch über die Fortschritte der Mathematik|JFM]]&nbsp;[http://www.zentralblatt-math.org/zmath/en/search/?q{{=}}an:{{{input1|}}}&format{{=}}complete {{#tag:nowiki|{{{input1|}}}}}]
      }}<!--
   -->{{only in print
       |JFM&nbsp;{{{input1|}}}
      }}
 |jstor={{hide in print
         |[[JSTOR]]&nbsp;[http://www.jstor.org/stable/{{{input1|}}} {{#tag:nowiki|{{{input1|}}}}}]
        }}<!--
     -->{{only in print
         |JSTOR&nbsp;{{{input1|}}}
         }}
 |isbn={{hide in print
        |[[International Standard Book Number|ISBN]]&nbsp;[[Special:BookSources/{{{input1|}}}|{{{input1|}}}]]
       }}<!--
    -->{{only in print
        |ISBN&nbsp;{{{input1|}}}
       }}
 |issn={{hide in print
        |[[International Standard Serial Number|ISSN]]&nbsp;[http://www.worldcat.org/issn/{{{input1|}}} {{#tag:nowiki|{{{input1|}}}}}]
       }}<!--
    -->{{only in print
        |ISSN&nbsp;{{{input1|}}}
       }}
 |mr={{hide in print
      |[[Mathematical Reviews|MR]][http://www.ams.org/mathscinet-getitem?mr{{=}}{{{input1|}}} {{#tag:nowiki|{{{input1|}}}}}]
     }}<!--
  -->{{only in print
      |MR{{{input1|}}}
     }}
 |oclc={{hide in print
        |[[Online Computer Library Center|OCLC]]&nbsp;[http://www.worldcat.org/oclc/{{{input1|}}} {{#tag:nowiki|{{{input1|}}}}}]
       }}<!--
    -->{{only in print
        |OCLC&nbsp;{{{input1|}}}
       }}
 |pmc={{hide in print
       |[[PubMed Central|PMC]]&nbsp;[http://www.pubmedcentral.gov/articlerender.fcgi?tool{{=}}pmcentrez&artid{{=}}{{{input1|}}} {{#tag:nowiki|{{{input1|}}}}}]
      }}<!--
   -->{{only in print
       |PMC&nbsp;{{{input1|}}}
      }}
 |pmid={{hide in print
        |[[PubMed Identifier|PMID]]&nbsp;[http://www.ncbi.nlm.nih.gov/pubmed/{{{input1|}}} {{#tag:nowiki|{{{input1|}}}}}]
       }}<!--
    -->{{only in print
        |PMID&nbsp;{{{input1|}}}
       }}
 |zbl={{hide in print
        |[[Zentralblatt MATH|ZBL]]&nbsp;[http://www.zentralblatt-math.org/zmath/en/search/?q{{=}}an:{{{input1|}}}&format{{=}}complete {{#tag:nowiki|{{{input1|}}}}}]
       }}<!--
    -->{{only in print
        |ZBL&nbsp;{{{input1|}}}
       }}
}}

Headbomb {talk / contribs / physics / books} 21:05, 3 May 2010 (UTC)

Done —TheDJ (talk • contribs) 22:59, 4 May 2010 (UTC)

Thanks a bunch! Headbomb {talk / contribs / physics / books} 23:14, 4 May 2010 (UTC)

Authorlinks in book citations

They seem to be glitching out and producing tags ("[" and "]") and pipes ("|") when giving authorlinks in book citations. Here's an example:

Monaghan, Patricia (January 2004), The Encyclopedia of Celtic Mythology and Folklore, Facts on File, p. 512, ISBN 978-0816045242 {{citation}}: Check |authorlink= value (help); External link in |authorlink= (help)

Anybody know how to fix this? :O Macai (talk) 12:22, 8 February 2010 (UTC)

I contrived a rather crude temporary fix until someone with administrative powers can edit the template to work with authorlinks in books. Basically, you add an external link that starts with the surname of the author, and you close that link with the end of the author's given name. Here's an example of how to do it:

Monaghan, Patricia (January 2004), The Encyclopedia of Celtic Mythology and Folklore, Facts on File, p. 512, ISBN 978-0816045242 {{citation}}: External link in |last= (help)

Hope this helps. Macai (talk) 12:27, 8 February 2010 (UTC)

Please read the documentation. Authorlink is for an internal link to the Wikipedia article on the author and is not intended for an external link. For example:

{{citation |title=[[Baden-Powell (book)|Baden-Powell]] |last=Jeal |first=Tim |authorlink=Tim Jeal |publisher=Yale University Press |year=2001 |isbn=0-300-09103-6}}

Jeal, Tim (2001), Baden-Powell, Yale University Press, ISBN 0-300-09103-6

---— Gadget850 (Ed) ^talk 13:13, 8 February 2010 (UTC)

Ohh. Okay. My apologies, then. Macai (talk) 13:22, 8 February 2010 (UTC)

No apologies needed— this question comes up every now and then. I took a look at the documentation and it could use some work. "Authorlink works either with author or with last & first to link to the appropriate article" is not very informative. ---— Gadget850 (Ed) ^talk 13:36, 8 February 2010 (UTC)

I've amended the documentation. --Redrose64 (talk) 15:56, 10 May 2010 (UTC)

Collapsible quote feature

Would there be any interest in collapsing the contents of the quote= field in a manner similar to the show/hide links we see on navboxes. (For example, a [show quote] handle following the citation (preceding the postscript?), in lieu of a fully-expanded quote.)

Sometimes the quotes can be a bit overwhelming when compared to the rest of the citation, so this might lessen the clutter. Since I suspect that the quote is primarily used to do a web search with a big block of relevant, unique text, most of the time, it doesn't need to be seen. For non-default situations, like when you're attempting to draw notice to a particular passage, a simple parameter showquote=yes could display it. TheFeds 07:48, 15 May 2010 (UTC)

Performance issues

Some of our longer articles such as Israel, United States and George W. Bush take a very long time to prepare on the server (i.e. almost a minute), and it appears it is because of the large number of citation templates. There is currently a discussion about this at WP:VPT#Israel's article. One option would be to replace all citation templates by hand-coded citations with the same appearance, but I wonder if there are less drastic options. Hans Adler 16:52, 10 May 2010 (UTC)

Such testing as has been carried out appears to indicate that the citation templates add very little indeed to the page download time. Malleus Fatuorum 17:22, 10 May 2010 (UTC)

I might suggest looking at why NewPP generates the following in the html:

"Post-expand include size: 1976735/2048000 bytes" for Israel, "Post-expand include size: 1904112/2048000 bytes" for United States" compare: "Post-expand include size: 835573/2048000 bytes" for Wikipedia" I'll also offer the observation that these articles are very heavy on images, if that has any bearing. Still, each of the three took less than 70ms to generate, I'm not sure where your minute is coming from. LeadSongDog come howl 17:49, 10 May 2010 (UTC)

My minute is coming from actual, reproducible experience. I just tried it again with Israel, and the following is at the end of the source code: "Served by srv201 in 28.190 secs." So it was only half as long, but I did have to wait for this half minute.

The same problem exists on a relatively short sandbox page that consists only of references. See my test results at WP:VPR#Setting an upper limit to the number of templates in given article. Hans Adler 18:46, 15 May 2010 (UTC)

last/first vs last1/first1

Please see new thread at cite journal and comment there. --Redrose64 (talk) 13:11, 20 May 2010 (UTC)

nopp

Done
In the subsection Parameters: nopp is listed as an argument, but is not explained. Should it be added? -DePiep (talk) 15:20, 17 April 2010 (UTC)

It suppresses the "p." or "pp." that normally appears before the page number for non-journal citations. In many cases |at= can be used instead. Consider this:

{{citation|title=A book|last=Doe|first=John|year=2010|page=section 12 para. 4}}

Doe, John (2010), A book, p. section 12 para. 4

It looks a bit odd, so you could add |nopp=yes to give

{{citation|title=A book|last=Doe|first=John|year=2010|page=section 12 para. 4|nopp=yes}}

Doe, John (2010), A book, section 12 para. 4 {{citation}}: Unknown parameter |nopp= ignored (|no-pp= suggested) (help)

but |at= is shorter:

{{citation|title=A book|last=Doe|first=John|year=2010|at=section 12 para. 4}}

Doe, John (2010), A book, section 12 para. 4

--Redrose64 (talk) 16:58, 17 April 2010 (UTC)

I know what it is for (e.g. it is explained in {{template:cite book}}-doc). The question is: should it not be in the list here. -DePiep (talk) 00:14, 18 April 2010 (UTC)

My take on this is that it should be explained at citation/doc. I've added an explanation. Wtmitchell (talk) (earlier Boracay Bill) 00:27, 18 April 2010 (UTC)

Is what I tried to ask, thx. May be expand it a bit there, with logic & relation to "page=", plus an example (as Redrose64 did here)? -DePiep (talk) 09:05, 18 April 2010 (UTC)

... the /doc is not protected, of course, so I could edit it myself. -DePiep (talk) 16:25, 22 April 2010 (UTC)

I wrote a template, called surprisingly {{Page numbers}} which will selectively add p. or pp. For example:

p. 4
pp. 4-15

The intent was for it to be built into the citation family to make them easier/better. It would need additional work to perform the above functions, currently it gives

Section 4

Rich Farmbrough, 18:01, 27 May 2010 (UTC).

Yes I couldn't resist improving it. Rich Farmbrough, 18:19, 27 May 2010 (UTC).

Parameter "at"

Parameter "at" appears to hev been changed to "loc", but the documentation has not been changed to reflect that. Could someone please update it? Malleus Fatuorum 12:24, 29 May 2010 (UTC)

Not done - it's still |at= in {{citation}} - the templates which have |loc= include {{harv}}, {{harvnb}} and {{sfn}}. --Redrose64 (talk) 12:42, 29 May 2010 (UTC)

You're right, my mistake. Malleus Fatuorum 13:25, 29 May 2010 (UTC)

Citations download with Zotero

I vaguely remember that one could download all the references cited on Wikipedia into Zotero as individual entries with a single click (at least all those made with templates). I could be mistaken though (http://forums.zotero.org/discussion/327/wikipedia/ http://www.zotero.org/blog/zotero-wikipedia-perfect-together/ ) but Zotero has changed and so have the citation templates and now it seems the citation download produces only the citation for the Wikipedia article page. Seems like a useful feature to have if confirmed it would be nice to have the required metadata in the new template. Shyamal (talk) 10:36, 30 May 2010 (UTC)

Look at the browser address bar. To the right of the URL you will see some icons. The folder icon should have a hover tip to Save to Zotero. ---— Gadget850 (Ed) ^talk 17:27, 31 May 2010 (UTC)

Thanks a lot. That was the feature I was looking for. Shyamal (talk) 03:13, 1 June 2010 (UTC)

New citation template possible??

Could I ask a favour from one of the very knowledgeable people out there on constructing a new citation template? I think it would be fantastic if it were possible to have a template that formatted references in the OSCOLA style (this is "Oxford Standard Citation of Legal Authorities). The main thing is its style (which I think is simple and attractive, and is used in the courts and most legal articles now) for books and articles. For instance to be complicated,

C Mitchell, 'Assistance' in P Birks, Breach of Trust (Hart Publishing 2002) ch 6, 209-211

... for a chapter in a book, or as a normal journal article:

S Gardner, 'Knowing Assistance and Knowing Receipt: Taking Stock' (1996) 112 Law Quarterly Review 56

More is on the OSCOLA page. But would it be possible to have a **citeoscola** entitled template? If anyone can help, please do let me know, cheers, Wikidea 16:34, 31 May 2010 (UTC)

Project Gutenberg

I am currently using this template to format Battle of Waterloo#References (I have done a couple in on the page so that I could discuss the concept on the talk page) but I have copied the section to my sandbox and will do the rest there before copying back all of them to the article page. I have worked my way down the list adding in the missing data for entries like the first one for Bonaparte. But I would like some advise on how best for format the next line using this template:

Fitchett, W. H. (1897, reprint 1921 & 2006). Deeds that Won the Empire. Historic Battle Scenes. London: John Murray. (Project Gutenberg). Chapter: King-making Waterloo

I looked at the documentation and searched the talk pages, but could not extrapolate from those the best way to format the above, keeping as much information as is practical, so some help would be appreciated. -- PBS (talk) 02:52, 3 June 2010 (UTC)

Like this:

* {{citation |last=Fitchett |first=W. H. |title=Deeds that Won the Empire. Historic Battle Scenes |year=1897, reprint 1921 & 2006 |publisher=John Murray |location=London |chapter=King-making Waterloo |url=http://www.gutenberg.org/etext/19255 |chapterurl=http://www.gutenberg.org/files/19255/19255-h/19255-h.htm#chap1900}}

Fitchett, W. H. (1897, reprint 1921 & 2006), "King-making Waterloo", Deeds that Won the Empire. Historic Battle Scenes, London: John Murray {{citation}}: Check date values in: |year= (help); External link in |chapterurl= (help); Unknown parameter |chapterurl= ignored (|chapter-url= suggested) (help)CS1 maint: year (link)

There is no need to reference Project Gutenberg as the source of the convenience link. ---— Gadget850 (Ed) ^talk 13:10, 3 June 2010 (UTC)

I'd like to mention that since {{citation}} emits COinS metadata, which includes the value of the |year= parameter, that parameter shouldn't really contain anything other than the year. An examination of the page source shows this:

rft.date=1897%2C+reprint+1921+%26+2006&

that is, something expecting a date is being passed a non-date value. I'd use these parameters instead:

* {{citation |last=Fitchett |first=W. H. |title=Deeds that Won the Empire. Historic Battle Scenes |year=1897 |edition=reprint 1921 & 2006 |publisher=John Murray |location=London |chapter=King-making Waterloo |url=http://www.gutenberg.org/etext/19255 |chapterurl=http://www.gutenberg.org/files/19255/19255-h/19255-h.htm#chap1900}}

Fitchett, W. H. (1897), "King-making Waterloo", Deeds that Won the Empire. Historic Battle Scenes (reprint 1921 & 2006 ed.), London: John Murray {{citation}}: External link in |chapterurl= (help); Unknown parameter |chapterurl= ignored (|chapter-url= suggested) (help)

--Redrose64 (talk) 13:30, 3 June 2010 (UTC)

Frack— I meant to split that. ---— Gadget850 (Ed) ^talk 13:46, 3 June 2010 (UTC)

Website parameter

Hello. There was a discussion held at Template_talk:Cite_web#.22Work.22_vs_.22Publisher.22_parameters that brought up the possibility of adding a website= parameter, which would greatly reduce (for some people like me) the confusion between the work= and publisher= parameters, as well as the auto-italicizing seen in the work= parameter. Any thoughts? – Kerαunoςcopia^◁_galaxies 06:29, 11 June 2010 (UTC)

ISBN Link Discussion

There's a proposal to make an alteration to so that a citation ISBN links is consistent with the magic ISBN mediawiki produces links. Please review and comment.

Lunchboxhero (talk) 13:27, 15 May 2010 (UTC)

Any updates on this? Helder (talk) 18:03, 19 June 2010 (UTC)

Wiki doi is a covert advertising instrument!

Hello. The way the Wiki DOI is currently operated violates one of the core principles of Wikipedia, namely not to become a vehicle of promoting or advertising external suppliers of goods or information. Imagine ISBN would be automatically linking books from Wikipedia to Amazon or any other online book store. That would be a gross violation of this principle, wouldn't it?

But exactly this is happening with the way doi is linking articles from Wikipedia to external sources: It promotes the publishing company Wiley, Project MUSE and JSTOR by linking directly and exclusively to their site, and I noticed that User:Citation bot—certainly unwilling, but nonetheless— has become its automatized instrument by which the whole process has been recently speedied up throughout WP.EN.

Take this example:

Wirsching, Armin (2000), "How the Obelisks Reached Rome: Evidence of Roman Double-Ships", The International Journal of Nautical Archaeology, 29 (2): 273–283, doi:10.1111/j.1095-9270.2000.tb01456.x

The citation template links to Wiley by default. But why Wiley? Actually ScienceDirect has also published this article! So why is Wiley picked over ScienceDirect? And, in fact, there are innumerable scholarly articles which have appeared simultaneously in a number different online data bases such as PAO, ProQuest, EBSCO, Free E−Journals, Ingenta, Extenza etc. Why should Wiley, MUSE and JSTOR be preferred over all these and many more as seems now the case?

My recommendation:

For the short term, stop immediately any action by User:Citation bot until this question is settled.
Take the way Wiki handles ISBN numbers as a model and create a meta page like Book sources which at least allows the reader to make an individual choice between the various data bases. This way Wikipedia would become much less of an effective advertiser of certain data bases over others.
Alternatively, link journal articles directely to the home page of the publishing house, in this case to www.nauticalarchaeologysociety.org.

Whatever we do, we need to draw line to not play advertising agent for anyone, even if some pay data bases like JSTOR don't consider themselves commercial. Gun Powder Ma (talk) 11:40, 11 June 2010 (UTC)

You're joking right? Headbomb {talk / contribs / physics / books} 11:51, 11 June 2010 (UTC)

Do all these data bases take substantial fees or not for access (20+ $)? They do, hence there is a strong commercial element involved and it is hard to see why Wikipedia should play marketing agent for any one of them or some of them to the detriment of the others. If this is not taking sides, then what is? We would never dream of linking directly to Amazon, either. ISBN is only for identification of the book as such, not for promotion of those who offer it for sale, and so should doi only be for identifyfing the journal article as such, but not by promoting these commercial data bases. Gun Powder Ma (talk) 12:08, 11 June 2010 (UTC)

Don't mock others Headbomb. It is unkind. And in this instance it makes you look arrogant and unwise (which I'm sure you're not). The problem is a real one. I have wondered about this for some time. We are linking our readers to one vendor preferentially over alternative vendors. It is advertising; very effective advertising, too, judging by how much Amazon pays to third parties for just such traffic. Anthony (talk) 12:10, 11 June 2010 (UTC)

So what is your suggested solution for citations, then? ╟─Treasury Tag►You may go away now.─╢ 12:12, 11 June 2010 (UTC)

No problem—when citing sources, naturally, their authors/publishers/vendors get a free advert. There's no way around this. And for what it's worth, temporarily stopping citation bot to fix such a non-problem (!) is an absurd idea. ╟─Treasury Tag►sundries─╢ 12:12, 11 June 2010 (UTC)
You can say that, just like that, can you TT? 30 seconds thinking, and the categorical answer is: "There's no way round that." Cool. Everybody go home. We're done here. Anthony (talk) 12:19, 11 June 2010 (UTC)
Nice little joke! :) ╟─Treasury Tag►most serene─╢ 12:21, 11 June 2010 (UTC)
I think you're trying to say there is no solution. There is a problem, right? Anthony (talk) 12:25, 11 June 2010 (UTC)
I don't really see an issue here. It is inevitable that, when citing sources, those who stand to profit from those sources' publicity will profit. I don't consider this majorly problematic, and the alternative (not citing sources) is clearly not viable. ╟─Treasury Tag►Not-content─╢ 12:27, 11 June 2010 (UTC)
The only alternative? Inevitable? Thanks. Anthony (talk) 12:29, 11 June 2010 (UTC)
(edit conflict) As I understand it, the problem boils down to the fact that when an article cites a fact to The History of the Oven Glove, Joe Bloggs, International Publishing Inc., 1990 then both Mr Bloggs and International Publishing Inc. get a bit of a boost. This also applies when citing a journal article available online: 18th Century Oven Glove Decoration, Rowena Bloggs, University of Delaware Press, 1978. Academic Journal Articles Online Inc.
As I see it, the only conceivable way around mentioning the names of sources, authors, vendors and publishers – which is the only "advertising" involved – is to not mention the names of sources, authors, vendors and publishers. If you have any alternative suggestions, I would welcome hearing them. ╟─Treasury Tag►estoppel─╢ 12:35, 11 June 2010 (UTC)
(edit conflict x3)I'm going to hold off on forming an opinion just yet but I must say that I'm leaning towards the view of Gun Powder Ma. I'm eager to hear opposing views on this. J04n(talk page) 12:13, 11 June 2010 (UTC)
The objection "we need to draw line to not play advertising agent for anyone, even if some pay data bases like JSTOR don't consider themselves commercial" is nonsense. doi does not prefer a source, it guarantees that some source will be available even if a previous one is unavailable. Payment for journal article pay data bases is no worse that having needing for someone to pay for a book you want to cite. --Philcha (talk) 12:33, 11 June 2010 (UTC)
(edit conflict) My point precisely. If giving the name of a source constitutes advertising, then, yep, we need to advertise to that extent. ╟─Treasury Tag►constabulary─╢ 12:37, 11 June 2010 (UTC)

The underlying purpose of DOI is to give a link to a canonical website established by the publisher who assigned the DOI. According to our article, "Referring to an online document by its DOI provides more stable linking than simply referring to it by its URL, because if its URL changes, the publisher need only update the metadata for the DOI to link to the new URL". Of course we want to link to an online version of the references we use, whenever possible. It's best to use DOI when we can, rather than an ad hoc link to the publisher's website, which is what we would use otherwise. If there is a free version or preprint online we should link to that as well, but a link to the canonical version is also important. So there's no problem here. — Carl (CBM · talk) 12:36, 11 June 2010 (UTC)

Thanks Carl. Sounds good to me. Anthony (talk) 12:40, 11 June 2010 (UTC)

Carl's view sounds about right. The {{citation}} and {{cite journal}} documentation already recommends that a free-access URL be provided for any citation where such a URL is available. Perhaps what's needed is a bot to add such links from these other databases, if they offer free access. Rjwilmsi 12:43, 11 June 2010 (UTC)

Carl, this issue goes far beyond rather mundane problems such as providing stable links. What Wikipedia, the largest free-content site on the whole internet, is actually doing here is linking tens of thousands journal articles, probably in the order of hundreds of thousands, to two to three commercial vendors - for free. If I were Wiley's manager, I would congratulate myself for each day the Citation bot adds another thousand links for free to my pay site.

No, this is a misrepresentation of the situation. The people who have linked their content to two or three commercial vendors are the academics who have published their papers in commercial journals. What we are doing in Wikipedia is merely providing accurate citations to those papers. Your argument (that we should avoid using DOI) is an argument that our citations should be less accurate, and I have no sympathy for that. —David Eppstein (talk) 15:54, 11 June 2010 (UTC)

(edit conflict)

Further to those, if no such is found, an issn or oclc link to journals and other serials will help readers and editors to find their nearest libraries which hold the work. Wikilinking the journal name will achieve the same effect if we have an article on the journal, as {{infobox journal}} has both |issn= and |oclc=. It's a clear case of wp:BTW. If the keeper is game, this would be a nice complement to the tasks that user:Citation bot handles. LeadSongDog come howl! 15:46, 11 June 2010 (UTC)

But doi should not be connected to any one commercial vendor any more than ISBN numbers are connected directly to vendors. Otherwise, we would establish a double standard, wouldn't we? We should work towards a solution where doi linking is as independent from commercial journal data bases as our ISBN linking is from commercial book stores. The current automatized linking to a few selected publishing houses completely flies in the face of Wikipedia:NOTADVERTISING. Gun Powder Ma (talk) 15:31, 11 June 2010 (UTC)

It doesn't. It connects to the Digital Object Identifier resolver system. That system is not driven by wp, or Wiley for that matter. You might want to actually read up on how it works.LeadSongDog come howl! 15:46, 11 June 2010 (UTC)

The reason that we don't link from an ISBN to a particular vendor is that many bookstores will sell the same book. In the case of journal articles, the only group that distributes Elsevier articles is Elsevier (via ScienceDirect), the only group that distributes Springer articles is Springer, and so on. There isn't a multiple-distributor market for reprints of journal papers like there is with books. This monopoly is a motivation behind open-access publishing, but there are very few open-access journals at the moment. The reality is that the majority of papers are only available in electronic form through a single website chosen by the publishers of the journal.

We should link to a free version or preprint of a paper, if one is available, of course. But we should also link to the canonical site for the publication. Even if our reader doesn't have free access to the body of the paper, she can usually obtain for free: (1) verification of the publication details of the paper; (2) an abstract of the paper; (3, in some cases) a list of references from the paper. Also, many of our readers come from academia, where they do have access to the bodies of articles through university-wide licenses. So the links to the canonical webpages are extremely useful even if not every reader can make full use of them. — Carl (CBM · talk) 15:51, 11 June 2010 (UTC)

Ditto on the preceding comments. Sure, access to journals is often (usually?) limited to subscribers, but we still need to access them, and DOI is very important for getting the authoritative details of publication. And most journals provide a free abstract of articles, so someone who does not have access to an actual copy can still get some sense of what the article covers and its worth when a free version is not available. - J. Johnson (JJ) (talk) 16:17, 11 June 2010 (UTC)

We should use care in linking free online versions of papers. Often they are preprints from an early point in the refereeing process, before the authors made some important corrections. The quality of the source should be a higher priority than its price. —David Eppstein (talk) 18:45, 11 June 2010 (UTC)

For sure. Which is why we should always (ideally) have a DOI to the the authoritative source. As to rating the quality of any convenience links to free copies, that's a different issue. - J. Johnson (JJ) (talk) 19:08, 11 June 2010 (UTC)

DOI has nothing to do with this

I agree with the responses above, but to return to the article cited by the original poster. This is not a problem with DOI, which will take you to the proper publisher, but this paper is carried by both Wiley Interscience and Elsevier's ScienceDirect because the journal transferred to Wiley-Blackwell in 2004 with its archive, but the previous publisher still carries the old papers. They both have DOIs assigned:

I guess one should make a general decision whether in these cases to cite the original publisher or the current owner. Vesal (talk) 00:31, 12 June 2010 (UTC)

I suppose one could also complain to the DOI foundation that this is not supposed to happen and ask them to please fix it. —David Eppstein (talk) 01:01, 12 June 2010 (UTC)

Or perhaps ask the copyright holder, the Nautical Archaeology Society? LeadSongDog come howl! 02:38, 12 June 2010 (UTC)

It's not illegal "advertising" to say that a book is published by XYZ publisher--it's a fact that readers need. Nor is it "advertising" to link to sites that millions of Wiki users have free access to, -- sources like JSTOR (the great majority of college students and the many millions of people who use libraries have free access.) These sites sell their services to libraries, not to individual Wikipedia users. Note that linking an article to, say, JSTOR, does not give any revenue to JSTOR or anyone else. Rjensen (talk) 18:25, 12 June 2010 (UTC)

JSTOR is a pay site to which not a single user has free access to (students pay student fees). There is no reason why we should treat commercial journal data bases a priori any different than commercial book stores. Both make money with what they do and neither provides free access. Gun Powder Ma (talk) 22:21, 12 June 2010 (UTC)

Wrong. In fact, anyone can walk into the library of a public university that subscribes to JSTOR, use the computers there, and access JSTOR. For free. Regardless of whether they are a student or have paid student fees of some sort. And in any case your argument is pointless: this is an encyclopedia of all human knowledge, not an encyclopedia of free stuff available on the net. Being for-pay does not invalidate a source or make it inappropriate for us to use. —David Eppstein (talk) 00:42, 13 June 2010 (UTC)

The "anyone can walk into the library..." argument is being undermined as schools start requiring accounts just to get on the computers. But that is really no different than needing an account to check out print materials. And the main point is, as David says, to build on ALL human knowledge, not just the free stuff on the Internet. To which end DOI is a necessary index. - J. Johnson (JJ) (talk) 19:20, 13 June 2010 (UTC)

Indeed. DOIs are the internationally recognised method for indexing and locating journal articles of all kinds. This is much ado about nothing, especially considering the only argument against is that they aren't all free, whereas multiple arguments have been presented for keeping them, not to mention they are the only real option at this point if you want these journal articles to be of any real value to the end user. — Huntster (t @ c) 19:39, 13 June 2010 (UTC)

Multiple-distributor market for journals

As I take it, the following two key assertions by J. Johnson (JJ) from above provide the rationale for why journal vendors are treated differently than book sellers. I will show below that both contentions cannot be uphold.

There isn't a multiple-distributor market for reprints of journal papers like there is with books...The reality is that the majority of papers are only available in electronic form through a single website chosen by the publishers of the journal.

[Carl (CBM · talk) is the actual source of this quote - J. Johnson (JJ) (talk) 19:35, 13 June 2010 (UTC)]

In fact, the whole market of online journal data bases is no less competitive and complex than that of online book stores. The following is only a tiny selection from the field of history, restricted to journals beginning mostly with A. I could add another hundred from my personal working list alone; there must be tens of thousand journal titles which are distributed by more than one vendor.

Acta Archaeologica, year 2003, is availabe at the following competing vendors:
- EBSCO
- Blackwell/Wiley

Agricultural History, year 1995, is availabe at the following competing vendors:
- PAO
- JSTOR

Agricultural History, year 2002, is availabe at the following competing vendors:
- ProQuest
- JSTOR

Annals of Science, years 1975-2004, is availabe at the following competing vendors:
- EBSCO
- Informaworld

Annual of the American Schools of Oriental Research, year 1922-1995, is availabe at the following competing vendors:
- JSTOR
- PAO

Antiquity, years 2001-2009, is availabe at the following competing vendors:
- EBSCO
- ProQuest

Antike und Abendland, years 1945-1989, 1992-2000, is availabe at the following competing vendors:
- PAO
- de Gruyter

Archive for History of Exact Sciences, years 1998-2009, is availabe at the following competing vendors:
- Springer
- EBSCO

Capitalism, Nature, Socialism, years 2003-2009, is availabe at the following competing vendors:

Der Islam, years 2004-2009, is availabe at the following competing vendors:

So why do we allow doi linking to pick commercial data bases like JSTOR consistently over all its other competitors? If this is not a case of promoting one enterprise to the detriment of the others, then what is? Gun Powder Ma (talk) 22:21, 12 June 2010 (UTC)

We're not going to stop using DOIs. It's an established standard and very widely used by people who read scholarly journals. The absence of a 'fairer' system to allow people to pick their vendor shouldn't mean we hobble ourselves by refusing to use DOIs out of some misplaced purism. Fences&Windows 22:51, 12 June 2010 (UTC)

So you concede the point that the Wikipedia doi linking system is currently favouring some vendors over others? Gun Powder Ma (talk) 22:57, 12 June 2010 (UTC)

It links to the official source, I don't see a problem with that. Fences&Windows 00:55, 13 June 2010 (UTC)

I wondered if anyone had ever criticised DOI in this way before, and I did find a note saying that some librarians complain about it directing to the publisher site as this can cause double purchasing - the library has access via one means, but the library user doesn't realise this and purchases it via the site they land on when they click the DOI.[1] There is an issue, but I don't think it's a solid justification for removing all DOIs from Wikipedia. Fences&Windows 01:41, 13 June 2010 (UTC)

Thanks for finding this source — I added it to our digital object identifier article. —David Eppstein (talk) 02:01, 13 June 2010 (UTC)

There is no "Wikipedia doi linking system" - We link to the standard DOI resolver. If you know of a "fairer" one, please suggest it. But not including DOIs because they don't link to all databases equally is not a good idea. It would be quite a step backward. As F&W stated, its a widely used standard in academia. Mr.Z-man 20:41, 13 June 2010 (UTC)

I have pointed out a fairer system in my initial post under My recommendation. In fact, I would say that almost any system which stops the systematic preferential treatment of a few selected journal databases would be more fair than what we have now. Gun Powder Ma (talk) 11:36, 17 June 2010 (UTC)

To go off on a flight of fancy for a moment, there is no reason, so far as I can see, that WMF couldn't run its own doi resolver. That resolver could be made to respond in accordance with priorities more suitable to foundational ideals. It could resolve to the most open, most recent, most authoritative, or some other "best" distributor. In the process, it might be possible to squeeze out some performance improvements given that the vast majority of dois are not referenced in any WP article. But I would not be prepared to bet that we could do this better than doi.org can without seeing some evidence.05:25, 14 June 2010 (UTC)

Regarding the first of two supposed assertions (mistakenly attributed to me) GPM provides evidence of a "multiple-distributor market for journals" in the field of history. However, I do not see the same evidence for scientific articles, such as appear in Science, or Geological Society of America Bulletin. Is this a counter-example? Or could similar evidence be shown for scientific journals?

As to the second supposed assertion: just what is it? - J. Johnson (JJ) (talk) 19:31, 14 June 2010 (UTC)

Replacing DOI with another resolver that finds open access versions of articles is not a trivial problem, e.g. [2][3]. OAIster and Google Scholar haven't cracked it, so I don't think we will. Fences&Windows 23:29, 15 June 2010 (UTC)

DOI is not the problem (2)

Wait .. DOI here is NOT the problem. DOI is here a good link, it links, stable, to the version an author used to write a sentence/statement on Wikipedia. So, the first example:

Wirsching, Armin (2000), "How the Obelisks Reached Rome: Evidence of Roman Double-Ships", The International Journal of Nautical Archaeology, 29 (2): 273–283, doi:10.1111/j.1095-9270.2000.tb01456.x

The citation template links to Wiley by default. But why Wiley? Actually ScienceDirect has also published this article!

Although I can see that the citation bot is here 'wrong', these are not often occuring cases. The template should be adapted to hold multiple DOI's. If this original template was not added by citationbot, but by a human editor, then the doi cited should be the doi that points to the version the author used.

But then, this is NOT advertising because there is a DOI, this is advertising because we reference. Even a citation:

Wirsching, Armin (2000), "How the Obelisks Reached Rome: Evidence of Roman Double-Ships", The International Journal of Nautical Archaeology, 29 (2): 273–283

is advertising. And what if this contains not any clickable link, but a set of pointers to writer names and or to journals is still there. With a slight adaptation:

Wirsching, Armin (2000), "How the Obelisks Reached Rome: Evidence of Roman Double-Ships", The International Journal of Nautical Archaeology, 29 (2): 273–283

the journal is still clickable.

Although refspamming and self citation would be a form of 'advertising', in the end we always end up linking to companies who actually earn money from the incoming traffic (or even, let us pay before we can even see it). The only solution would be, to remove all references to commercial entities .. and I don't think that is what we want. But the problem is not the doi (except for the cases where there are multiple dois possible, where we should consider to link them all, or none of them). --Dirk Beetstra ^{T C} 08:04, 17 June 2010 (UTC)

Note: almost any form of external link in a reference can serve a beneficial effect on the organisation linked to. Also many not-for-profit organisations need money from 'beneficiaries' (being it musea from incoming 'real' visitors (who saw the existence of the museum via a Wikipedia reference), or libraries from government grants). Incoming internet visitors, even when not recorded where they are from, do show up in the 'server load' of the organisation linked to (Google hits and Google ranking is not the only measure of success). Having links from Wikipedia will, even when added in good faith by many editors who have no involvement at all with the organisation, show that your organisation is efficient, that your information is sought after, and that you are doing your job as an organisation well (and hence, could be a measure to show you need more money). So, in the end, our readers (and editors) are sponsoring said organisation, even while totally unintended. The point hence is, agian, the doi itself is not the problem, the links would be. Having dois and other links is just a price we pay for trying to get rid of this (the good thing is, some of these organisations gladly link back to Wikipedia, despite this, which probably sponsors Wikipedia again). --Dirk Beetstra ^{T C} 08:22, 17 June 2010 (UTC)

I believe the argument against the current selective DOI linking is a straightforward one. To name the journal as such is not advertising by any common sense, but to link to them (even selectively!) the way it is currently done is. The advertising problem lies with the linking which gives quick and exclusive access to a certain journal supplier to the exclusion of its competitors. We don't link books via ISBN directly to online book stores for good reasons, so by analogous reasoning we should not link articles via DOI directly to online journal vendors, either. Gun Powder Ma (talk) 11:33, 17 June 2010 (UTC)

Wowow, that is a step too quick. Yes, some sources are available from multiple organisations, but by far the most is only available from one. And whether or not it is a clickable link, it is still advertising at the moment that someone references the Wiley copy (because they used thát copy) and not another one. --Dirk Beetstra ^{T C} 11:36, 17 June 2010 (UTC)

Referring to a source is not advertising, but Wikipedia:Citing sources. But linking, thousand upon thousand times, to a specific commercial vendors who charges fees falls nothing short of Wikipedia:PROMOTION. As for the quantitaive dimension: How many more examples of articles available from multiple organisations would you like to see to modify your judgment? I am serious, give me a realistic number and I will post them in time here. As I said, I know of many dozens in the field of history and archaeology alone. Gun Powder Ma (talk) 11:42, 17 June 2010 (UTC)

How many examples can you provide of articles from the Geological Society of America Bulletin being available from somewhere other than the GSA site? (There are some, but usually where an author has grabbed one of his own articles to post on his personal page, and in many cases these are only preprints, and not authoritative.) Can you show any of the GSAB articles cited at, say, Olympic-Wallowa Lineament#References as "available from multiple organisations"? - J. Johnson (JJ) (talk) 20:49, 20 June 2010 (UTC)

The thing is, I am from a different field (chemistry), and I don't see that happen too often. If something gets published by, say, ACS, then there will not be a copy of the same document on a Wiley server. I have seen cases where there are two essentially the same publications in two different journals of two different publishers, but exactly the same, no. But I can see that that is for certain information not the case. Here I see the same article in the same journal, but the journal is apparently published via (at least) two different publishers, each of them assigning an own DOI to it.

A good thing is, I hope this will be random, one time the DOI will point to one, the other time to the other (I mean, one time the doi on wiki). But I don't think it is a good idea not to link. Maybe CitationBot should be made aware of this issue. --Dirk Beetstra ^{T C} 12:00, 17 June 2010 (UTC)

It's odd because this is a violation of the DOI spec. But I doubt the publishers would be interested in hearing about it. — Carl (CBM · talk) 13:13, 17 June 2010 (UTC)

In a couple of days, I'll present a number of scholarly articles which appear at least in two journals; I am under the impression that some users still prefer to ignore the double standard which comes by the current selective doi linking system. Gun Powder Ma (talk) 13:02, 22 June 2010 (UTC)

According to the NAS website that journal is published for them by Wiley-Blackwell. They participate in CrossRef which freely serves bibliographic metadata for use by any user, including by Citation bot. The former publisher of that journal, Elsevier, use a proprietary for-fee database, Scopus instead. This puts them in a COI with a disincentive to keeping the DOI current. Are you advocating that WMF should pay Elsevier to use that database in order to find their outdated DOIs? LeadSongDog come howl! 13:36, 22 June 2010 (UTC)

For DOIs of citations that you add, yourself, this is a non-issue: WP:SAYWHEREYOUGOTIT (just as you would for a book with multiple ISBNs). I agree there could be some ambiguity for bot-added DOIs. The utility of having any DOI out-weighs the concern over "fairness:" what we are looking at is a mere corner case. --Karnesky (talk) 14:58, 22 June 2010 (UTC)

I call bullshit on this statement that "some users still prefer to ignore the double standard which comes by the current selective doi linking system." It postulates this alleged "double standard" which you have not demonstrated (outside of one possibly anomalous area). It is not that the rest of us "prefer to ignore" what you think is so blatant, but that you have not shown that there is any "double standard" or such to accept. Also, I object to your characterizing the opposition here as "some users". In the small set of "users" (editors?) attending this discussion it appears to me that the opposition is total, that your position has no support. - J. Johnson (JJ) (talk) 22:12, 22 June 2010 (UTC)

Poll

I think the discussion, as long discussions tend to do, begins to blur what is in essence a clear issue. Therefore, I would like the participants to answer a few questions of mine:

Do you agree that Wikipedia in spirit and Wikipedia:PROMOTION in letter forbids the systematic promotion of commercial, non-free information suppliers?
Do you agree that online journal databases such as JSTOR, MUSE, PAO, EBSCO, ProQuest etc. are commercial digital libraries which restrict access to the paying customer?
Do you agree that doi, as it is currently implemented in Wikipedia, systematically and directly links to such vendors by automatized means such as User:Citation bot? Do you further agree that this linking prefers some commercial databases to the exclusion of others?

Please provide yes or no answers, although additional comments are welcomed, too, of course. Gun Powder Ma (talk) 11:33, 17 June 2010 (UTC)

No, as it is not (proven) to be systematic promotion of one over the other; I agree, but it does not matter, the question is verifyability; No, I do not believe that Citation bot is preferentially programmed to use one over another. And whether linked or not, when the document is uniquely on one vendors website, we do already promote that vendor (and the bigger publishers hence get more incoming traffic from Wikipedia than other publishers). --Dirk Beetstra ^{T C} 12:23, 17 June 2010 (UTC)

Since you ask ...

No. There are no such words in WP:PROMOTION.
Not really. That category's inclusion criteria twist the normal meaning of "commercial". Few if any academics would consider a university, public, or government library as a "commercial" enterprise. JSTOR merely federates their access on a non-profit basis. Including the German Project Gutenberg in the cat is simply bizarre.
No. Linking to the resolver operated by doi.org is in no sense a preferential selection of "vendors", it is quite blind to whether the target is published on commercial terms or even whether there is a cost-recovery fee.
LeadSongDog come howl! 13:53, 17 June 2010 (UTC)

My answers

Yes, if it's done for commercial interests, which isn't the case here. That the DOI link to commercial sites is irrelevant: DOIs are the official links, stay up to date even after journals merge, split, change publishers, etc., guarantees that we are not linking to an illegal copy of the material, allows for bots to find missing citation information, etc...
Yes, but it's irrelevant.
The first part is trivially factual, the second part is irrelevant.
Headbomb {talk / contribs / physics / books} 14:00, 17 June 2010 (UTC)

This whole poll seems to be based on a logical fallacy to me. Possibly the fallacy of the undistributed middle, I'm not sure. What I am sure of is that whatever answers I give are going to be twisted to support your quixotic anti-referencing campaign. I refuse to play that game. —David Eppstein (talk) 04:24, 19 June 2010 (UTC)

The current DOI linking system isn't done for commercial interests, it uses the official system. That aside, I'd like to bring up the issue of linking to google books over amazon book preview, both of which offer limited book previews on certain books.Smallman12q (talk) 20:24, 20 June 2010 (UTC)

This whole poll is specious. Any time something is forced into an "either/or" format it generally leaves out much and damages the rest, leaving the results wholly unreliable. In this case Gun Powder Ma has taken a narrow interpretation ("systematic promotion of commercial, non-free information suppliers") and crossed it with another ("JSTOR, MUSE, PAO, EBSCO, ProQuest etc. are commercial digital libraries which restrict access...") as if they are fully conformable – but they are not. I could more fully explain this, but, sorry, I have other things to do. I would be more interested if the approach taken had been "is this a valid syllogism?" (which we might discuss), but presented as "if you accept this you must accept that" rather puts me off. - J. Johnson (JJ) (talk) 21:07, 20 June 2010 (UTC)

authormask extension

Could please someone extend the authormask= so that by entering author2mask= one can substitute the 2nd author's name with em dashes (and author3mask= for the third, and so on)? Just like the author-link= parameter. --bender235 (talk) 13:02, 5 July 2010 (UTC)

Anyone? --bender235 (talk) 11:39, 9 July 2010 (UTC)

"This parameter is primarily intended for use in bibliographies where multiple works by a single author are listed." Can you give an example where the mask for two authors would be useful? ---— Gadget850 (Ed) ^talk 13:58, 9 July 2010 (UTC)

Not for two authors, but for the second author (or third, ...). For example, in the bibliographie of Peter Grünberg, there is an entry

G. Binasch, P. Grünberg, F. Saurenbach, W. Zinn. 1989. "Enhanced magnetoresistance in Fe-Cr layered structures with antiferromagnetic interlayer exchange". Physical Review B39. 4282.

that should displayed via {{Citation}} as

G. Binasch, ———, F. Saurenbach, W. Zinn. 1989. "Enhanced magnetoresistance in Fe-Cr layered structures with antiferromagnetic interlayer exchange". Physical Review B39. 4282. --bender235 (talk) 10:31, 10 July 2010 (UTC)

I don't see that as useful, a potentially confusing. Perhaps if all the authors were the same, but not just the one. I wouldn't even use the mask if the first author was the same, but the second different. Is there a style guide that shows the usage you desire? ---— Gadget850 (Ed) ^talk 11:20, 10 July 2010 (UTC)

I don't know. But what is your alternative? --bender235 (talk) 13:54, 11 July 2010 (UTC)

To use the full names without a mask. Using dashes in this manner appears to be Chicago style, where this template is APA style. Not sure why it is a feature. ---— Gadget850 (Ed) ^talk 14:02, 11 July 2010 (UTC)

It also appears that this should only be used where the list is formatted with hanging indentions. {{Reflist}} does not support indentions, but {{Refbegin}} does}} (which reminds me to update the documentation to not use hanging indentions and columns together). ---— Gadget850 (Ed) ^talk 16:42, 11 July 2010 (UTC)

page and pages

pages: For use when multiple pages are cited. Adds "pp." before the page numbers. Do not use with page.
page: For use when one page is cited. Adds "p." before the page number. Do not use with pages.

The text says the above, but it does not add the pp. or p. The examples on the page also show this. Snowman (talk) 22:32, 3 August 2010 (UTC)

It does add the pp. and p. for chapters in books. It does not add them for journal articles, because the standard formatting for journal articles is different from that for book chapters. —David Eppstein (talk) 00:31, 4 August 2010 (UTC)

If so, I think that it should say that with the rest of the explanation to reduce confusion. Snowman (talk) 11:48, 4 August 2010 (UTC)

Translation parameters

See Template talk:Citation/Archive 3#Translator, Template talk:Citation/Archive 3#Translator 2 and Template talk:Citation/Archive 3#trans_title

Translation parameters would be useful. -- PBS (talk) 04:22, 3 June 2010 (UTC)

The last discussion that you linked ended with a four-point request, the first of which was for the required changes to be sandboxed. This has not occurred; the sum total of sandbox changes since that moment are here. --Redrose64 (talk) 12:38, 3 June 2010 (UTC)

For citing historical sources it would be very useful to have parameters for entering the translators' names; in the contexts of translated works the translators often play the role that editors do for edited texts. The coding (which is beyond me) shouldn't be that different than the corresponding author and editor parameters and I request that a knowledgeable editor to consider making such changes.

For historical uses, it is not normal to cite the original title for published translated works; I imagine people citing translated editions of modern works would want to cite the original title as well as the English translation. From my perspective, there doesn't seem to be quite as urgent a need for this change. --SteveMcCluskey (talk) 04:14, 5 August 2010 (UTC)

If all that you need is the means to add the translator's name, the |others= field is available for this, except when citing patents. Use eg |others=trans. Doe, John --Redrose64 (talk) 14:15, 5 August 2010 (UTC)

parameters for parenthetical notes at end

It's sometimes important to add information to a citation that is not provided for with a citation template. An example is an author's credential. U.S. legal writing employs these often, such as to distinguish like-named cases decided in the same court and year or to clarify the reason for citing a particular case or state its adjudicatory stage.

My present solution is to write the citation in the ref element without a template, so, if the template is present, I remove it.

You might want to add a parameter for a parenthetical end-of-citation note. More than one such parameter per citation should be acceptable. Probably the order of appearance of such parenthetical notes should be controllable by the citation editor.

This is for Citation and Cite templates, probably all of them.

Thank you. Nick Levinson (talk) 17:30, 8 August 2010 (UTC)

Most of the templates have the |postscript= parameter, which may be used for this purpose; however, for some templates such as {{cite web}}, it doesn't work if |quote= is also present. --Redrose64 (talk) 18:09, 8 August 2010 (UTC)

The official explanation isn't even close to supporting that. Can someone who knows the parameter better than I do please edit the page? As far as I can tell, the parameter is only for adding punctuation, not for enclosing a statement.

Here's a hypothetical use case:

<ref>Jones v. Smith, 142 N.Y.2d 353 (1977) (railroad liability).</ref>

I don't know how to add "(railroad liability)" if I were to use a Citation or Cite template, with or without a postscript parameter.

And we'd want the template to work even if a quotation is also present.

Thank you. Nick Levinson (talk) 19:14, 8 August 2010 (UTC) Corrected example's markup: Nick Levinson (talk) 19:23, 8 August 2010 (UTC)

You can just type the text after the citation template. This is one reason why the template's output does not end with a period.

Author, Title {{citation}}: |author= has generic name (help), extra information

{{citation|author=Author|title=Title}}, extra information

— Carl (CBM · talk) 23:24, 8 August 2010 (UTC)

NBSP and hyphens in citations

There is a discussion at User talk:Art LaPella#Your AWB edits concerning whether WP:NBSP should be applied within date parameters of a citation template as in date={{Nowrap|6 November}} 2010. It also concerns whether hyphens within titles should be changed to dashes according to the WP:DASH rules that apply elsewhere, as in this previous discussion. LeadSongDog come howl! 15:24, 19 August 2010 (UTC)

location field for citing a periodical when article may be print online only?

I have been involved in a discussion with Parrot of Doom (talk · contribs · count) over the use of |location= in citations. If the citation is to an article on the website of a periodical (in this case a UK daily newspaper, if that matters), and it's not known whether or not the article was included in a printed edition of the periodical, are there particular rules on the use of |location=. I had understood that the appropriate citation would be:

{{citation|...(URL, author, date, accessdate)...|work=The Guardian | location=London }}

whereas Parrot of Doom (talk · contribs · count) says the correct format is

{{citation|...(URL, author, date, accessdate)...|publisher=guardian.co.uk }}

I've used |location= on the basis of the documentation wording: location: Place of publication for journal or periodical. This is probably not a big deal, but is using |location= the right approach? Thanks Rjwilmsi 14:40, 25 August 2010 (UTC)

Your question seems not to be what the section header implies, but the reverse.

Our usual practice, at least recently, has been pretty lax about |place= and |publication-place=, but they should really be used. For a newspaper article, the best practice should be to base the place parameter on the article's dateline and base the publication-place on the paper's masthead. For an online-only news source, because domain names are very perishable, I would prefer to see

{{citation |author=Roving Reporter |title=A Royal getaway |date=25 August 2010 |place=[[Holyrood Palace]] |work=[[The Guardian]] |publication-place=London |edition=Online |url=http://guardian.co.uk/blahblahblah |accessdate=25 August 2010 }}

to render as

Roving Reporter (25 August 2010), written at Holyrood Palace, "A Royal getaway", The Guardian (Online ed.), London, retrieved 25 August 2010

so that it is clear to the reader where it was written, where it was published, and where you got it. LeadSongDog come howl! 16:38, 25 August 2010 (UTC)

Yes, title was wrong, I've amended it. Rjwilmsi 17:01, 25 August 2010 (UTC)

Agency, newspaper, and location

This question seems to be related to the question above, at least somewhat.

Currently, if {{Cite news}} is given an agency, a newspaper, and a location, it will list them in the following order:

Newspaper. Agency (Location).

However, it would make more sense to list the location immediately following the newspaper -- the location refers to the where the newspaper is published, not where the agency is located. Where the agency goes in this case is less of a priority to me, as long as it is put somewhere logical. But the other two items should be displayed in this way:

Newspaper (Location).

I raised this issue at Template talk:Cite news and was told that this was dependent on {{citation/core}}, so then I brought it up at Template talk:Citation/core, where nobody has responded yet. Since this talk page appears to be better read, I am raising the issue here. Please note that I am only concerned here about the display of {{Cite news}}, and I don't know how {{Citation}} relates to that. Maybe it doesn't, but if both templates use {{citation/core}}, perhaps the {{citation/core}} experts will see this post here. --Metropolitan90 (talk) 18:08, 25 August 2010 (UTC)

Agencies are something of a thorny problem. While some are pure syndication distributors, others have staff writers too. A source crediting "AP Staff" should be used with author=Associated Press staff or similar. The |agency= parameter is useful when AP merely takes content from one paper and syndicates it to others. It seems clear that the syndicates add little if any fact checking, so where possible I would argue that we should try to identify the originating publication and source article content from that paper's version if possible. For current events, google news' date ranking option can be helpful, though not authoritative, in finding out the original publisher. LeadSongDog come howl! 19:48, 25 August 2010 (UTC)

(edit conflict) Please see WP:MULTI - it's best to start a discussion in one place, not three (this does not prevent you posting a note here which draws attention to the original). The original discussion at Template talk:Cite news#Agency, newspaper, and location seems to have generated firther comment. --Redrose64 (talk) 19:53, 25 August 2010 (UTC)

ISBN Links

{{editprotected}} The ISBN links produced by this template are different from the ISBN links produced by Mediawiki's ISBN Magic. Specifically, the ISBN links produced by this template lack the class name of the magic ISBN links.

ISBN 0952545705

Special:BookSources/0952545705

produces

<a href="/wiki/Special:BookSources/0952545705" class="internal mw-magiclink-isbn">ISBN 0952545705</a>

<a href="/wiki/Special:BookSources/0952545705" title="Special:BookSources/0952545705">Special:BookSources/0952545705</a>

This inconsistency makes it difficult to internationalize this popular script that rewrites ISBN links.

Proposal

replace

|isbn={{hide in print
        |[[International Standard Book Number|ISBN]]&nbsp;[[Special:BookSources/{{{input1|}}}|{{{input1|}}}]]

with

|isbn={{hide in print
        |[[International Standard Book Number|ISBN]]&nbsp;ISBN {{{input1|}}}|{{{input1|}}}]]

Lunchboxhero (talk) 01:51, 13 May 2010 (UTC)

That would show up as "ISBN ISBN 0123456789012]]" Headbomb {talk / contribs / physics / books} 06:01, 13 May 2010 (UTC)

However, replacing

 |isbn={{hide in print
        |[[International Standard Book Number|ISBN]]&nbsp;[[Special:BookSources/{{{input1|}}}|{{{input1|}}}]]
       }}<!--
    -->{{only in print
        |ISBN&nbsp;{{{input1|}}}
       }}

with

 |isbn=ISBN&nbsp;{{{input1|}}}

should work. I have no particular opinion on whether or not the change should be made. Headbomb {talk / contribs / physics / books} 06:06, 13 May 2010 (UTC)

Not done for now: Sorry I am not comopletely satisfied that this is an uncontroversial change and this talk page of a subtemplate is unlikely to receive much notice. Would you mind discussing the merits of this proposal somewhere appropriate first? — Martin (MSGJ · talk) 05:39, 14 May 2010 (UTC)

I would support this. Where should we talk about this request if this isn't the right place? Helder (talk) 20:16, 19 May 2010 (UTC)

I put a notice of this proposal on the talk page for the Citation template ten days ago, and there has been no response. Lunchboxhero (talk) 16:16, 25 May 2010 (UTC)

The code should be simply

 |isbn=ISBN {{{input1|}}}

Even the non-breaking space is overkill- the wiki-renderer must do the right thing. Rich Farmbrough, 22:04, 27 May 2010 (UTC).

The nbsp isn't overkill at all, and ensures that ISBN 0123456789012 will display on a single line, like all other identifiers (ISSN, doi, bibcode, JSTOR, etc...). Headbomb {talk / contribs / physics / books} 17:39, 9 July 2010 (UTC)

Done — Martin (MSGJ · talk) 11:19, 10 July 2010 (UTC)

Putting an ISBN into a {{citation}} or {{cite book}} used to make the ISBN linkable, but now, it doesn't. It doesn't matter whether the ISBN is 10-digit or 13; whether it is formatted with spaces, hyphens or none - none of these are linkable. The code

{{citation/identifier |identifier=isbn |input1=0-7153-6769-2 }}

should generate

ISBN 0-7153-6769-2

instead it generates

ISBN 0-7153-6769-2

see the difference? So, {{editprotected}} please revert {{citation/identifier}} to the previous version and check again. --Redrose64 (talk) 15:41, 10 July 2010 (UTC)

Done Reverted back to version of 22:58, 4 May 2010 Ronhjones ^(Talk) 20:04, 10 July 2010 (UTC)

Has consensus been established about what behavior is desired? FWICS, handcoded cites containing e.g., ISBN 1234567890 don't wrap as a unit. I think that there's a question (discussed a bit above, but not driven to consensus-supported decision) re whether or not cites using citation/core should wrap like such handcoded cites. Wtmitchell (talk) (earlier Boracay Bill) 00:52, 13 July 2010 (UTC)

I think it makes sense that the ISBN link behave like the other identifiers, and not line break. Lunchboxhero (talk) 02:59, 13 July 2010 (UTC)

I wouldn't disagree with that, but it raises the question in my mind of whether the identifiers should behave differently with templated cites than ISBN behaves with hand-crafted cites. I don't have a strong view about that, but it looks to me like a valid question. If the answer is "No", the handling of the other identifiers should change. Wtmitchell (talk) (earlier Boracay Bill) 03:56, 13 July 2010 (UTC)

Why are we worried about whether the line breaks inside the link? It happens for wikilinks and urls all the time. As more and more usage is on small (mobile) screens this gets increasingly normalized. LeadSongDog come howl! 16:11, 13 July 2010 (UTC)

I'm amenable either way. Should we invite Headbomb to add input, or are we ready to move forward? Lunchboxhero (talk) 02:26, 15 July 2010 (UTC)

Continued resolution(s)

I think the problem was the nonbreaking space in the template:

|isbn=ISBN&nbsp;{{{input1|}}}

. As Redrose pointed out

ISBN 0-7153-6769-2

generates ISBN 0-7153-6769-2, while

ISBN&nbsp;0-7153-6769-2

generates ISBN 0-7153-6769-2. It seems that there are three options at present:

Ditch the nonbreaking space, in which case we cannot ensure that the identifier will appear on one line, as pointed out by Headbomb.
Ask the developers to alter the way the parser deals with ISBN numbers, so that they will honor nonbreaking spaces between "ISBN" and the ISBN number.
Accept that the ISBN links produced by this template are different than the ISBN links the parser produces.

Is there a way to specify the class of the ISBN link? Is there a way to ensure that the ISBN link does not line break without using

&nbsp;

?

Lunchboxhero (talk) 03:18, 11 July 2010 (UTC)

Should be able to wrap it in <span style="white-space:nowrap;">...</span>. ---— Gadget850 (Ed) ^talk 11:41, 11 July 2010 (UTC)

Also, is there a way to test any changes without delinking ISBN numbers in Notes and References sections throughout Wikipedia?

— Paine (Ellsworth's Climax) 11:34, 12 July 2010 (UTC)

Use the sandbox and testcases links at the bottom of the documentation. ---— Gadget850 (Ed) ^talk 12:14, 12 July 2010 (UTC)

<span style="white-space:nowrap;">...</span> doesn't seem to be accepted, but the

{{nowrap}}

template does. See Template:Citation/identifier/sandbox and Template:Citation/testcases. Lunchboxhero (talk) 02:59, 13 July 2010 (UTC)

I'm not yet sure why, but the testcases don't seem to fit automatically into the available window width, they keep a constant width instead, so that they have to be horizontally scrolled to read them. Definitely not good behaviour. LeadSongDog come howl! 15:23, 19 August 2010 (UTC)

LeadSongDog, I'm not sure what you are talking about. Could you explain more? Lunchboxhero (talk) 01:05, 26 August 2010 (UTC)

See the rendered example below. Note the lines don't wrap inside the box, but that there is a horizontal scrollbox at the bottom instead.

LeadSongDog come howl! 20:13, 30 August 2010 (UTC)

Format of names when citing patents

The formatting of the list of authors when citing books and journals appears to be very different from that obtained for the list of inventors when citing patents. As an example, the syntax:

{{citation | inventor1-last=Caro | inventor1-first=H. | inventor2-last=Kern | inventor2-first=A. | title=Manufacture of dye-stuff | issue-date=1883 | patent-number=290856 | country-code=US }}

gives:

Manufacture of dye-stuff {{citation}}: Unknown parameter |country-code= ignored (help); Unknown parameter |inventor1-first= ignored (help); Unknown parameter |inventor1-last= ignored (help); Unknown parameter |inventor2-first= ignored (help); Unknown parameter |inventor2-last= ignored (help); Unknown parameter |issue-date= ignored (help); Unknown parameter |patent-number= ignored (help)

while I would have expected:

Caro, H.; Kern, A., "Manufacture of dye-stuff", US patent 290856, issued 1883

Why is this? Is there an easy way to change this behaviour? Aa77zz (talk) 14:41, 31 July 2010 (UTC)

This expectation appears to be based on a misunderstanding of patent citation methods. See my post below. GDallimore (Talk) 12:52, 30 August 2010 (UTC)

Citing patents

I'm afraid that whoever has been updating this template to cite patents and then forcing the other patent citation templates to refer to it appears not to know much about how patents are cited. An assumption has been made that patents should be cited like other publications, but this assumption is simply not correct.

The main problem is that inventors are almost NEVER important and almost never cited. They are certainly not comparable with authors. The inventor field is pretty much the LEAST important field in the template, not the most important. The title is also typically not important, although more important than the inventor names. The most important fields in the citation are the two letter country code and the patent number. If you want proof of this, just look at how the links are generated: with the country code and patent number. That is how patents are located, identified and referenced.

This template needs reworking to take account of the above facts in the following ways.

The order of the output needs to be put to "country code" "patent number" "title". The remaining parameters can take pretty much any order, but I would put the inventors at or near the end.

The anchor should be based on a concatenation of the country code and patent number. That way there will always be an anchor, and it will be unique.

While making these changes, I think the "issue date" parameter should be changed to "grant date". "Issue" is US-only terminology. "Grant" is used worldwide.

Thanks. GDallimore (Talk) 12:50, 30 August 2010 (UTC)

Can you give us a reference that illustrates a proper style? ---— Gadget850 (Ed) ^talk 21:02, 30 August 2010 (UTC)

(commenting from the sidelines) I did some googling and found the following, which seem to disagree with the above:

How to Cite a Patent - examples for various citation styles
IEEE Citation Style Guide - page 2
How to Cite References using AIAA Format
International Committee of Medical Journal Editors (ICMJE) Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Sample References
References/Bibliography Harvard Style - page 8
How To Cite References - Vancouver Style

The above is not comprehensive. Wtmitchell (talk) (earlier Boracay Bill) 22:41, 30 August 2010 (UTC)

(edit conflict) Hi. There's not much wrong with how it's being cited at the moment, it's just the order of the output that needs changing to the one suggested above. I created this template a while ago {{Cite patent}} which I think is pretty sensible in its basic ordering of the major elements. I'm sure there are improvements that could be made to the output of the less important elements such as publication/grant dates, inventors and assignee names, status etc. But the major elements of country, patent number and title are definitely the first three to list in a citation. GDallimore (Talk) 22:43, 30 August 2010 (UTC)

(resolving edit conflict) As for the link provided above: the IEEE is the most illuminating as to why citing the inventor is completely wrong. The IEEE suggests inventors are comparable with authors. That is totally and utterly wrong. Inventors almost never author their own patent specifications.

The best guide to use for citing patents is surely one from an organisation that knows what they're doing, eg a patent office.

Don't know if this link will work, but it should go to a search report citing relevant patent documents prepared by the United States Patent and Trademark office. They use a slightly different format to the one I suggest, but the key point is it's the country and patent number that're used to identify a patent document.

The full citation style used by the USPTO is: "Country" "number" (First inventor) "Publication date"

If you were to look at a search report from the European Patent Office (who place less importance on the inventor than the USPTO), you would find a slightly different citation style of: "country" "number" (assignee) "Publication date". Eg, see the search report for EP 2219406 (click on "original document and go to page 17).

I still suggest using the title in Wikipedia citations because that can be helpful for an understanding of the subject of the patent in question. That's rarely important for patent office search reports since the reader will know that the subject of the cited patent will be similar to the subject of the patent it is being cited against. GDallimore (Talk) 23:08, 30 August 2010 (UTC)

(a end dot pls)

Cite web, cite book and cite journal include a dot at the end. Citation template does not. Could it be included too, pls. --Chris.urs-o (talk) 11:33, 11 June 2010 (UTC)

I'd rather like to see the redundant end dot removed in all templates, since a dot only marks a full sentence (subject, verb, [object]) which all these templates, however, evidently are not. Gun Powder Ma (talk) 11:40, 11 June 2010 (UTC)

All you need do is add the parameter |postscript=. (note the dot which follows the equals sign). --Redrose64 (talk) 12:35, 11 June 2010 (UTC)

Why do we use commas to separate the various parts of the citation (source, date, accessdate, etc), and then finish without the period? That's pretty inconsistent to me. Adding the postscript option sounds like a backwards kind of solution. 95.209.91.231 (talk) 14:59, 12 September 2010 (UTC)

Because it lets you string extra things onto the end of the citation by putting them immediately after the template with a comma in between. That would not work if the period was automatically generated. — Carl (CBM · talk) 18:51, 12 September 2010 (UTC)

Absolutely. Please do not add any punctuation after the template. — Cheers, JackLee ^–talk– 19:44, 12 September 2010 (UTC)

Incorrectly rendered anchor tags

Hi, I'm having an issue with incorrectly rendered anchor tags in the article Maria Anna Mozart. For all works listed under 'Works of literature with Maria Anna Mozart as a main character', anchor tags generated under 'ref=harv' in the cite tags results in a year of 2010 instead of the 'date=xxx' field within the tag. For example,

*{{Cite book
| publisher = Alcina Press
| isbn = 9780955071300
| last = Bauld
| first = Alison
| title = Mozart's Sister
| location = [[London, UK|London]]
| date = 2005
| url = http://books.google.com/books?id=2jdsopOT4GIC
| oclc = 64769121
| ref = harv
}}

generates an anchor tag of #CITEREFBault2010 (...view selection source) instead of #CITEREFBault2005. Explanations? --Blehfu (talk) 17:16, 26 September 2010 (UTC)

You need to use year= instead of date=. —David Eppstein (talk) 17:54, 26 September 2010 (UTC)

Exactly. This is a known problem which I have mentioned before. The puzzling thing is that sometimes a |date= containing only a bare year is correctly interpreted as a year, and sometimes it's fouled up.

I may be wrong, but I believe that the problem only occurs for bare years which may also be interpreted as times (times without dates are treated as if they were times on today's date). That is to say, |date=1960 to |date=1999 should normally work as intended (interpreted as years), but |date=1900 to |date=1959 and |date=2000 to |date=2059 will probably be interpreted as if they were 19:00 14 November 2024 etc. Thus, your example |date=2005 is treated as 20:05 14 November 2024, and the template parser extracts the year from that, which is 2024. --Redrose64 (talk) 20:49, 26 September 2010 (UTC)

Bewildering and funny at the same time. Thanks for the tips. Blehfu (talk) 00:06, 27 September 2010 (UTC)

I've added logic to AWB to automatically clean up |date=2005 etc. to |year=2005, just as a cosmetic tidy up, but if the incorrect field use causes a functional problem I can make a bot request to clean up those affected: is it where the year is between 00 and 60 then? Rjwilmsi 13:22, 27 September 2010 (UTC)

Not entirely sure. I only did a quick test, see User:Redrose64/Sandbox7. --Redrose64 (talk) 17:47, 27 September 2010 (UTC)

general style question

I've noticed that some articles/bibliographies, particularly associated with Wikipedia:WikiProject_Mesoamerica use this:

{{refbegin|indent=yes}}<!--BEGIN biblio format. If indent param. is used,  please use a colon (:) instead of asterisk (*) for bullet markers in the references list -->
:
{{refend}}<!-- END biblio format style -->

This is the style I'm more accustomed to when writing papers, as opposed to the bullet format used on Wikipedia most frequently. Any thoughts on the matter?

I've also noted with disappointment, as was mentioned in Talk Archive 3, that the use of {{aut|xxx}} breaks Zotero. Shame, it looks kinda nice. :) (ex. Mesoamerica) Blehfu (talk) 05:20, 28 September 2010 (UTC)

citing multiple quotes from same source?

I've got cases where I want to refer to multiple sections of the same source. I don't want to write, e.g.,

<ref>
{{cite manual
 |     author = IBM
 |       title = IBM System/370 System Summary
 |         id = GA22-7001-6
 |    version = Seventh Edition
 |       date = December 1976
 |       page = 69
 |        url = http://www.bitsavers.org/pdf/ibm/370/systemSummary/GA22-7001-6_370_System_Summary_Dec76.pdf
 |      quote = this is from one place
 |  separator = ,
 }}
</ref>
<ref>
{{cite manual
 |     author = IBM
 |       title = IBM System/370 System Summary
 |         id = GA22-7001-6
 |    version = Seventh Edition
 |       date = December 1976
 |       page = 666
 |        url = http://www.bitsavers.org/pdf/ibm/370/systemSummary/GA22-7001-6_370_System_Summary_Dec76.pdf
 |      quote = and this is from another
 |  separator = ,
 }}
</ref>

To get the references

IBM (December 1976). IBM System/370 System Summary (PDF). Seventh Edition. p. 69. GA22-7001-6. this is from one place {{cite book}}: Unknown parameter |separator= ignored (help)
IBM (December 1976). IBM System/370 System Summary (PDF). Seventh Edition. p. 666. GA22-7001-6. and this is from another {{cite book}}: Unknown parameter |separator= ignored (help)

but would rather have something like

IBM (December 1976). IBM System/370 System Summary (PDF). Seventh Edition. p. 69. GA22-7001-6. this is from one place {{cite book}}: Unknown parameter |separator= ignored (help)
GA22-7001-6, System/370 System Summary, p. 666, "and this is from another"

Where the styles for rendering, e.g., quotes, page numbers, would be consistent.

Is there a supported way to do this using the {{cite}} templates? Shmuel (Seymour J.) Metz Username:Chatul (talk) 16:35, 26 September 2010 (UTC)

I like to use cite templates in the references section in conjunction with harvnb templates marked inline with subsequently generated footnotes in a notes section. Take a look at the article Maria Anna Mozart or Maurice Ravel to see it in action.

In your situation, I would take

{cite manual
 |     author = IBM
 |       title = IBM System/370 System Summary
 |         id = GA22-7001-6
 |    version = Seventh Edition
 |       date = December 1976
 |        url = http://www.bitsavers.org/pdf/ibm/370/systemSummary/GA22-7001-6_370_System_Summary_Dec76.pdf
 |      quote = and this is from another
 |  separator = ,
 |        ref = harv
 }}

then place <ref>{{harvnb|IBM|1976|p=whateverpagenumbergoeshere}}</ref> inline in the article. Then add a ==Notes== {{reflist}}to precede ==References==. Caveat: this is not the only way to do it, just my way. --Blehfu (talk) 17:10, 26 September 2010 (UTC)

Those articles use shortened footnotes. ---— Gadget850 (Ed) ^talk 21:52, 26 September 2010 (UTC)

I can't use {{harvnb|IBM|1976|p=whateverpagenumbergoeshere}} because there are multiple manuals with the same date. Even if I could, the {{harvnb}} documentation doesn't show quote= as valid. Shmuel (Seymour J.) Metz Username:Chatul (talk) 23:59, 26 September 2010 (UTC)

I think for harvnb they suggest using 1976a and 1976b for multiple sources that have the same year. If quotes don't go inline, they could go in the notes: <ref>{{harvnb|1976b|p=666}}. Quote goes here.</ref> Let me know what article you're working on currently and I can take a look. Blehfu (talk) 00:13, 27 September 2010 (UTC)

Using the year for citations of manuals has several issues; I'm looking for something that will allow me to use the vendors' order numbers as the identifier. I do put quotations in notes, but via a {{cite manual}} template in the footnote, and want each of the quotes subject to the same formatting.

The articles that I'm currently concerned with are

Input/Output Configuration Program in progress
Input/Output Control System in review
System Generation (OS) in review
OS/VS2 (SVS) in review
User:Chatul/IBM System/360 architecture in progress

Once I add the necessary reference material to User:Chatul/IBM System/360 architecture and put it up for review, I expect to significantly expand Count Key Data.

Thanks. Shmuel (Seymour J.) Metz Username:Chatul (talk) 12:12, 27 September 2010 (UTC)

Hmmm... I think I understand this, but let me explain in case I misunderstand. Identifying manuals by years is not workable for you because vendors might issue several manual revisions by year; Suffixing 'a', 'b', etc. to the year is less than useful because it is not meaningful. Given this, I might suggest something like {{Harvnb|IBM|January 1987|Ref=SA22-7085-1}}, rendering: IBM & January 1987 or (messier without an update to the {{Harvxxx}} docs) , coupled with something like: {{Harvnb|IBM|SA22-7085-1|Ref=SA22-7085-1}}, rendering: IBM & SA22-7085-1

{{cite manual | author = IBM | title = IBM System/370 Extended Architecture Principles of Operation | id = SA22-7085-1 | version = Second Edition | date = January 1987 | separator = , | postscript = | ref=SA22-7085-1 }}, rendering:
IBM (January 1987). IBM System/370 Extended Architecture Principles of Operation. Second Edition. SA22-7085-1. {{cite book}}: Unknown parameter |separator= ignored (help)

Wtmitchell (talk) (earlier Boracay Bill) 13:04, 27 September 2010 (UTC)

Just a small point: you could use "ref = CITEREFIBMSA22-7085-1" in the cite template which negates the necessity for "ref=SA22-7085-1" in the harvnb template. Blehfu (talk) 18:24, 27 September 2010 (UTC)

You can also use {{SfnRef}} aka harvid; see the documentation. ---— Gadget850 (Ed) ^talk 16:36, 28 September 2010 (UTC)

Blah blah blah blah blah!^[1] Ho ho ho ho ho ho ho!^[2]

Notes

^ IBM & SA22-7085-1, p. 12345. A Møøse once bit my sister ...
^ IBM & SA22-7085-1, pp. 1234567–12345678. No realli! She was Karving her initials on the møøse ...

References

IBM (January 1987). IBM System/370 Extended Architecture Principles of Operation. Second Edition. SA22-7085-1. {{cite book}}: Unknown parameter |separator= ignored (help)

well, that solves half of the problem, but it still leaves the question of causing a quotation to be rendered in a fashion consistent with the rendering of a quotation by a transcludeded {{cite}} template.

Thanks. Shmuel (Seymour J.) Metz Username:Chatul (talk) 19:20, 27 September 2010 (UTC)

Couldn't these be appended to the inline citation? Blehfu (talk) 03:04, 28 September 2010 (UTC)

Blah blah<ref>{{harvnb|blah|123|p=123}}, {{q|quote goes here}}</ref>

Are the {{cite}} templates guarantied to use {{q}}? Shmuel (Seymour J.) Metz Username:Chatul (talk) 09:35, 28 September 2010 (UTC)

{{cite}} deliberately omits any sort of punctuation following it to allow for addenda. The use of {{q}} or just straight quote marks is purely to your own taste. Blehfu (talk) 15:37, 28 September 2010 (UTC)

Note that {{q}} uses typographic quotes which are recommended against by WP:MOSQUOTE and will not match the straight quotes used by citation templates. ---— Gadget850 (Ed) ^talk 16:36, 28 September 2010 (UTC)

{{cite manual}} adds quotation marks around the text in quote=. I'm trying to find an equivalent of ibid and op. cit. using the Wikipedia templates. Shmuel (Seymour J.) Metz Username:Chatul (talk) 20:21, 28 September 2010 (UTC)

The original question was how to format references like:

IBM (December 1976). IBM System/370 System Summary (PDF). Seventh Edition. p. 69. GA22-7001-6. this is from one place {{cite book}}: Unknown parameter |separator= ignored (help)
GA22-7001-6, System/370 System Summary, p. 666, "and this is from another"

I took a quick look, but don't see any precedent for this style in any reference system. Chicago does allow the use of dashes for multiple references by the same author, but the rest of the citation is still in full.---— Gadget850 (Ed) ^talk 16:47, 28 September 2010 (UTC)

I've seen some references to ibid and op. cit. being out of fashion, but didn't see any helpful suggestions for alternatives. I was hoping that the issue had been addressed by the template developers so that I could automatically get a consistent style. Shmuel (Seymour J.) Metz Username:Chatul (talk) 20:21, 28 September 2010 (UTC)

I just found this link, Wikipedia:Footnotes#Style_recommendations, as well as the following template:

Guess it's not au gôut du jour in Wikipedia.. Blehfu (talk) 03:27, 29 September 2010 (UTC)

I tend to use the Shortened footnotes system. However, when the article already has an established ref style, it's not a good idea to alter that without good reason, and preferably consensus. So, in cases such as this, it's possible to combine them: use long inline refs for the first instance of a fact from a given book, together with a shortened footnote for subsequent refs.

The first statement.<ref>{{cite manual |author=IBM |title=IBM System/370 System Summary |id=GA22-7001-6 |ref={{harvid|IBM System/370 System Summary|1976}} |version=Seventh Edition |date=December 1976 |page=69 |url=http://www.bitsavers.org/pdf/ibm/370/systemSummary/GA22-7001-6_370_System_Summary_Dec76.pdf |quote=this is from one place |separator=,}}</ref> A second statement.<ref>{{harvnb|IBM System/370 System Summary|1976|p=666}}, "and this is from another"</ref>

Points to note are that the |ref= parameter has a {{harvid}} on it; and the parameters to that {{harvid}} are identical to those on the {{harvnb}}, except that the page number is omitted from {{harvid}}. --Redrose64 (talk) 17:55, 29 September 2010 (UTC)

I might have to settle for that, but I was hoping for some way to use {{cite manual|quote=foo}} while ensuring that subsequent quotations from the same source were formated the same as the first. Shmuel (Seymour J.) Metz Username:Chatul (talk) 20:42, 29 September 2010 (UTC)

To do that, you'd want a quote= parameter added to Harvnb and other templates in the {{Harv}} family, and you'd probably want to create a new template which all of those, along with {{Citation/core}} would use to format the quote in a consistent manner (perhaps put it at {{Citation/core/quote}}??). It'd be the work of an hour or so to implement that, probably less, but it'd take more than that to get consensus on it. It might be reasonable to approach that by posting a notice at Template talk:Citation/core that a change to that effect is planned unless there is objection, waiting a few days, and then doing it. I don't think that it'd be unreasonable to just boldly make the changes to the Harv... templates, as it'd be adding a new optional parameter, not changing what's already there. Wtmitchell (talk) (earlier Boracay Bill) 03:32, 30 September 2010 (UTC)

Re named references:

Named references are used when there are several cases of repetition of exactly the same reference, including the page number for books; they should not be used to cite different pages in the same book.

Also Template:Refref is being considered for deletion. Shmuel (Seymour J.) Metz Username:Chatul (talk) 20:42, 29 September 2010 (UTC)

Use of accessdates in external links

There is a conversation regarding the use of accessdates in the External links section here Wikipedia talk:External links#Question about accessdates. Your comments would be greatly appreciated. --Kumioko (talk) 13:46, 7 October 2010 (UTC)

Lower cased "retrieved"

Please fix the template(s) so that lower cased "retrieved" is rendered upper cased "Retrieved" in the articles after the save for consistency. All is One (talk) 22:24, 27 October 2010 (UTC)

Capitalization depends upon chosen punctuation and the template seems to handle this correctly.:

Brin, Sergey; Page, Larry, Google, retrieved 2010-10-27
Brin, Sergey; Page, Larry, Google, retrieved 2010-10-27 {{citation}}: Unknown parameter |separator= ignored (help)

--Karnesky (talk) 00:06, 28 October 2010 (UTC)

Right. After a comma, it should not be capitalized. — Carl (CBM · talk) 02:25, 28 October 2010 (UTC)

Thank you for the clarifications. All is One (talk) 20:13, 28 October 2010 (UTC)

Quote parameter used for irrelevant text

Is there a good basis for the quote parameter being wide open, or is its use limited to supporting article text?

There are very brief instructions for the quote parameter; only that the text be relevant. I am seeing an editor edit warring to introduce text through the quote parameter, text that could be incorporated into the article but which does not directly support the relevant part of the article where the cite sits. I would rather see this text brought into the article in a way that helps tell the story, or left out altogether. I feel the quote parameter could be used to shoehorn promotional or negative text in a POV manner, without the constraint of neutral writing and good reading flow in the article. Binksternet (talk) 13:05, 29 October 2010 (UTC)

I think that discussion of what should and what should not be quoted in a citation belongs at Wikipedia Talk:CITE, not here. I think that the documentation for the quote= parameter here should read something like, "text to be included in the citation, bracketed between double-quote marks". Wtmitchell (talk) (earlier Boracay Bill) 23:20, 29 October 2010 (UTC)

New ProveIt referencing tool

I'd like to inform people of the new ProveIt reference management tool developed by the ELC lab at Georgia Tech. It's designed to provide a convenient GUI for viewing, adding, editing, and inserting footnote references. It currently has full support for the Cite family of templates, and view, edit, and insert support for Citation. Superm401 - Talk 19:51, 8 December 2010 (UTC)

2010a and 2010b: need template fix or solution

Abstract: 5 digits work well in harvnb. But whole DATE conflict with YEAR. If DATE exist it hide YEAR (linking). So you should state BOTH such {citation|...|year=2010b|date=2010/06/23}.

Hello, while {harvnb} allow it, it seems that the Citation template don't allow years with 5 digits.

Frugal engineering is nice (Sehgal, Dehoff & Panneer 2010a, p. 2 harvnb error: multiple targets (3×): CITEREFSehgalDehoffPanneer2010a (help)) But is complex (Sehgal, Dehoff & Panneer 2010b, p. 2). References (years are not displayed!) Sehgal; Dehoff; Panneer (May 7), Frugal Engineering - powerful approach to develop products and services, The Economic Times `{{citation}}`: Check date values in: `\|date=` and `\|year=` / `\|date=` mismatch (help) Sehgal; Dehoff; Panneer (May 25), The Importance of Frugal Engineering (Detailed article) `{{citation}}`: Check date values in: `\|date=` and `\|year=` / `\|date=` mismatch (help); Cite has empty unknown parameter: `\|1=` (help)

Can this be overcome ? Yug (talk) 06:07, 19 December 2010 (UTC)

Oh, I found a way. DATE and YEAR are conflicting. That work fine when we delete the parameter |date= . If include DATE, then YEAR is broken

* {{citation|last1=Sehgal|last2=Dehoff|last3=Panneer|year=2010a|date=May 7|title=Frugal Engineering|url=http://www.onelink.com }}
* Sehgal; Dehoff; Panneer (May 7), Frugal Engineering {{citation}}: Check date values in: |date= and |year= / |date= mismatch (help)

Only YEAR, so YEAR is displayed:

* {{citation|last1=Sehgal|last2=Dehoff|last3=Panneer|year=2010a|title=Frugal Engineering|url=http://www.onelink.com }}
* Sehgal; Dehoff; Panneer (2010a), Frugal Engineering

Yug (talk) 06:11, 19 December 2010 (UTC)

The date field is only designed for a full date, not just a month day or day month. In these situations include |year=2010a and |date=May 7, 2010 and it should all work. Rjwilmsi 08:48, 19 December 2010 (UTC)

I have found the date and year parameters in {{citation}} so full of pratfalls that when I need to add a/b suffixes to a year I just build an explicit CITREF (e.g., "ref=CITEREFBlakelyothers2009a"). {{Harv}} works fine, and is insulated from complications in {{citation}}.

I don't to what extent date and year are actually broken, or just inadequately documented. Better documentation is certainly need. - J. Johnson (JJ) (talk) 20:06, 19 December 2010 (UTC)

Perhaps more robust would be to use |ref={{Harvid|Blakely|others|2009a}}, but I'd guess that the hidden internal convention for building the ID using the CITEREF prefix is unlikely to change. Wtmitchell (talk) (earlier Boracay Bill) 19:41, 20 December 2010 (UTC)

Thanks to Redrose64 for editing my comment above to fix my inadvertant use of =Ref instead of =ref as a parameter name in the example I gave. The confusion grows out of the Harv family using =Ref for the name of the parameter there which pairs with =ref here and in the {{cite xxx}} templates. I've made and had to fix this error elsewhere in the past, and likely will do so again in the future. Wtmitchell (talk) (earlier Boracay Bill) 22:41, 20 December 2010 (UTC)

Getting back to the original topic of this section, of course just adding the suffix (e.g. |year=2001a) would both display the suffix in the cite and incorporate it into the ID. Alternatively, an optional parameter could be added here and in {{Citation/core}} (perhaps =yearsuffix and =Yearsuffix) to build the suffix into the ID (either with or without displaying it in the cite) if that parameter is provided. Wtmitchell (talk) (earlier Boracay Bill) 23:03, 20 December 2010 (UTC)

One solution which I have used several times before is to have |date= as the full date, without any letter suffix on the year, and to also give |year= with the letter suffix; it is the latter that is used in the {{harv}}, {{harvnb}} or {{sfn}}, like this:

A statement {{harv|Doe|2010a|p=12}} Another statement {{harv|Doe|2010b|p=34}}

{{citation |last=Doe |first=John |date=May 7, 2010 |year=2010a |title=Doe's first paper this year }}

{{citation |last=Doe |first=John |date=May 25, 2010 |year=2010b |title=Doe's second paper this year }}

This produces:

A statement (Doe 2010a, p. 12) Another statement (Doe 2010b, p. 34)

Doe, John (May 7, 2010), Doe's first paper this year{{citation}}: CS1 maint: date and year (link)

Doe, John (May 25, 2010), Doe's second paper this year{{citation}}: CS1 maint: date and year (link)

The year is thus given twice, in different forms: one for display, the other for linking. Try it and see. --Redrose64 (talk) 13:21, 21 December 2010 (UTC)

laws

Hello, I was wondering if there is a special template for citing legal texts, I just did not now how to cite an ordinance at State Bank of Vietnam —Preceding unsigned comment added by 87.231.158.21 (talk) 11:01, 22 December 2010 (UTC)

Have a look at "Category:Law citation templates", but I don't think will be a specific template for citing Vietnamese legislation. You may be able to find a generic template for legislation, though. — Cheers, JackLee ^–talk– 14:26, 22 December 2010 (UTC)

Thank you, there is no template for Vietnam but I'll know where to look next time. —Preceding unsigned comment added by 87.231.158.21 (talk) 21:55, 22 December 2010 (UTC)

Discussion on standardization

There's a proposal to standardize citation styles in WIkipedia that may be of interest to users of this template, at Wikipedia_talk:Citing_sources/example_style. —David Eppstein (talk) 16:27, 11 January 2011 (UTC)

Documentations for Template:Citation/identifier and Template:Citation/authors

{{editprotected}} Could you add documentations to {{Citation/identifier}} and {{Citation/authors}}? Or at least empty documentations to maintain interwikis? --DixonD (talk) 11:28, 11 January 2011 (UTC)

Are interwikis really needed for subtemplates? Surely a link to other language versions of the main template is sufficient. — Martin (MSGJ · talk) 15:02, 11 January 2011 (UTC)

I disagree with you. Of course subtemplates can have different names on different wikis and it can be hard to find corresponding template in another language without interwikis. --DixonD (talk) 08:40, 12 January 2011 (UTC)

Yes, subtemplates may have different names on different wikis but I cannot think of a reason why you might want to go directly from Template:Citation/authors to es:Plantilla:Obra citada/autores. It seems unnecessarily bureaucratic to put categories and interwikis on subtemplates, for little benefit. Typically the subtemplates are listed on the main template's documentation. — Martin (MSGJ · talk) 10:50, 12 January 2011 (UTC)

The reason is simple - it makes porting (and updating afterward) of templates from en-wiki to some other wiki simplier. But of course it is "bureaucratic" for users from en-wiki, why should they care about wikipedia in other languages? --DixonD (talk) 16:39, 13 January 2011 (UTC)

Not done: please establish a consensus for this alteration before using the {{edit protected}} template. Logan _Talk ^{Contributions} 06:04, 13 January 2011 (UTC)

accessdate format

Hello,

I recently noticed that other citation templates generate capital letter when accessdate parameter is used (i.e. Retrieved 12 May 1998), whereas this template generates small letter (i.e. retrieved 12 May 1998). Is it possible fix this inconsistency in this template?1exec1 (talk) 18:55, 29 January 2011 (UTC)

That's not a bug, it's a feature. This is correct behavior. Other citation templates separate parts of the citation by periods; this one uses commas. It is correct to capitalize the next word after a period, and not to capitalize the next word after a comma. —David Eppstein (talk) 19:07, 29 January 2011 (UTC)

Which means you should not mix {{citation}} with {{cite web}} and the like, as they are different styles. See User:Gadget850/Citation templates for a comparison (need to move that into wikispace somewhere so others can beat it up). ---— Gadget850 (Ed) ^talk 19:09, 29 January 2011 (UTC)

Something that's missing from your comparison: {{citation}} works by default with the {{harv}} series of templates. With {{cite journal}} etc, one needs to include ref=harv to make it work with those templates; otherwise, they generate broken wikilinks. (Also a bug not a feature: somehow the editors of cite journal decided that the broken wikilinks were less critical than the issue that if you have multiple cites with the same author/year combination you will get multiple locations with the same html label.) —David Eppstein (talk) 19:49, 29 January 2011 (UTC)

See User:Gadget850/Citation templates— anchors. ---— Gadget850 (Ed) ^talk 02:29, 30 January 2011 (UTC)

Ok, thanks for explanation. 1exec1 (talk) 19:35, 29 January 2011 (UTC)

Openlibrary

Might it be possible to use openlibrary codes to link to works, such as this book? That might negate any linkrot, and would also reduce the length of an article slightly. Parrot of Doom 15:45, 2 February 2011 (UTC)

Overlinking

Does anyone know how to delink the term "ISBN" in the citation template? It's repeated over and over in blue in the references sections, which adds a lot to the visual clutter. SlimVirgin ^{TALK|CONTRIBS} 03:12, 13 February 2011 (UTC)

It's not a feature specific to {{citation}} - the same happens with {{cite book}} etc. The linking is done in {{citation/identifier}}, where such linking is conditional: but it's only non-linked for print copies, it's always linked for screen display, and there are no switches or similar which might turn off the linking. If you want an enhancement, the talk page for that subtemplate redirects back here, so Template talk:Citation/core might be the best place. --Redrose64 (talk) 11:59, 13 February 2011 (UTC)

Though the use of a citation template versus a hand formatted cite doesn't create any more ISBN links since the ISBN magic word links too. Rjwilmsi 19:31, 13 February 2011 (UTC)

I think the ISBN magic word links to Special:Booksources: ISBN 0-000-00000-0 rather than linking to ISBN. But I think that there's nothing strictly wrong with linking to "ISBN" in a citation. It creates a nice tooltip when you hover over the word; the same happens for DOI with the cite templates. If we could get the tooltip without the link, that would also be OK with me. — Carl (CBM · talk) 20:03, 13 February 2011 (UTC)

Yes, I see, link to ISBN and booksources versus just booksources. Though only a slight difference as the ISBN magic word still has ISBN in blue, so is as much blue as the citation template |isbn=, so I'm not sure how SlimVirgin thinks this is more cluttered? Rjwilmsi 20:32, 13 February 2011 (UTC)

Right - a manually formatted citation would have the same amount of blue even if no templates were used. — Carl (CBM · talk) 20:49, 13 February 2011 (UTC)

ZBL → Zbl

In {{citation/identifier}}, all instances of "ZBL" should be replaced with "Zbl". This is a mistake I made a long while ago, but somehow got overlooked (see Template talk:Citation/Archive 3#Identifier sandbox) Headbomb {talk / contribs / physics / books} 02:35, 16 February 2011 (UTC)

Done —David Eppstein (talk) 06:41, 16 February 2011 (UTC)

Edit request

Thread moved from Template talk:Cite because Template:Cite is now a redirect to Template:Citation --Redrose64 (talk) 12:36, 3 March 2011 (UTC)

{{editprotected}}

The |language field appears before the title instead of after it like the related templates {{cite web}}, etc. could someone please fix this?

e.g:

Someone, An example of the error ({{cite}})

Vs

Someone. "An example of the error". {{cite web}}: Missing or empty |url= (help) ({{cite web}})

Thanks, --Jeremy (blah blah • I did it!) 02:27, 2 March 2011 (UTC)

Not done for now: Sorry to be a pain but please could you follow the standard procedure:

Proposal
Discussion
Consensus
Sandbox
{{editprotected}}

A few steps in the middle have been missed out. Thanks — Martin (MSGJ · talk) 08:23, 2 March 2011 (UTC)

You're not being a pain, I saw something I thought was incorrect, and I asked for it to be fixed. There appears to be a reason for what I thought to be the inconsistencies and I can accept why they are there.--Jeremy (blah blah • I did it!) 19:22, 2 March 2011 (UTC)

I have moved the above thread from Template talk:Cite because Template:Cite is now a redirect to Template:Citation (this occurred over two years ago), therefore the discussion should be on this page. --Redrose64 (talk) 12:36, 3 March 2011 (UTC)

As replied to your comment at Wikipedia talk:Citing sources#Citation template inconsistencies, {{citation}} and {{cite book}} are two different styles. ---— Gadget850 (Ed) ^talk 12:47, 3 March 2011 (UTC)

type= parameter

This is noncontroversial — I think.

I've just sandboxed addition of support for a type= parameter. that parameter is supported by {{cite book}}, but is currently not supported here. I re-synch'd citation/doc/sandbox with the current docs and added documentation there for the added parameter.

Any objection to going live with this? 04:51, 8 March 2011 (UTC)

And I may have been asleep (and/or ignorant?), so — what is this supposed to do? - J. Johnson (JJ) (talk) 23:51, 8 March 2011 (UTC)

It's supposed to regularize

{{cite book|title=title|type=type}}: title (type).

more closely with

{{citation|title=title|type=type}}: title (type)

like this

{{citation/sandbox|title=title|type=type}}: title (type)

I don't see any reason for having the disparity. Wtmitchell (talk) (earlier Boracay Bill) 01:55, 9 March 2011 (UTC)

Also see Template talk:Citation/core/Archive 9#Proposal to add a "type" parameter. --Redrose64 (talk) 20:40, 9 March 2011 (UTC)

The related parameter in /core is TitleType=. Note that I recently changed the documentation there regarding the usage of that parameter. That change and a similar change to the {{Cite book}} docs grew out of discussion at Wikipedia talk:Citation templates#Audiobooks, and the documentation change I'll make here dovetails with those doc changes. Wtmitchell (talk) (earlier Boracay Bill) 00:16, 11 March 2011 (UTC)

Having seen no objection, I've gone live with the change. Wtmitchell (talk) (earlier Boracay Bill) 00:16, 11 March 2011 (UTC)

{{Citation}} use raised at WP:WQA

Re a long-term dispute with an editor on aircraft articles who strongly dislikes {{citation}}
Wikipedia:Wikiquette_alerts#user:Bzuk_and_the_removal_of_citation_templates

Andy Dingley (talk) 11:14, 14 March 2011 (UTC)

Many things about identifiers

Current behaviour

Currently, several identifiers are supported by {{citation}} and the {{cite xxx}} family.

John Smith (2000), How to make things out of other things, p. 123, Bibcode:0123456789, doi:10.0123456789, ISBN 0123456789, ISSN 0123456789, JSTOR 0123456789, OCLC 0123456789, PMC 0123456789, PMID 0123456789, ID0123456789 {{citation}}: Check |bibcode= length (help); Check |doi= value (help); Check |issn= value (help); Check |pmc= value (help); Check |pmid= value (help)

Order and JSTOR

It seems to me that there are no good reasons to prioritize any of them in their current order. In addition, the JSTOR parameter does not produce a JSTOR123456789 link at the end of the citation like others. This is not the biggest of problems, but it still should be present at the end regardless (it would make references in PDFs much nicer). Aka, once everything is said and done, I think citations should look like

John Smith (2000), How to make things out of other things, p. 123, Bibcode:0123456789Check bibcode: length (help), doi:10.0123456789, ISBN 0123456789, ISSN 0123456789 Parameter error in {{issn}}: Invalid ISSN., JSTOR 0123456789, OCLC 0123456789, PMC 0123456789, PMID 0123456789, ID0123456789

Print and JSTOR

Also, in the print version of the template, the JSTOR link should be suppressed

John Smith (2000), How to make things out of other things, p. 123, Bibcode: 0123456789, doi:0123456789, ISBN 0123456789, ISSN 0123456789, JSTOR 0123456789, OCLC 0123456789, PMC 0123456789, PMID 0123456789, ID0123456789

Otherwise, it'll be expanded into

John Smith (2000), How to make things out of other things (http://www.jstor.org/stable/0123456798), p. 123, Bibcode: 0123456789, doi:0123456789, ISBN 0123456789, ISSN 0123456789, JSTOR 0123456789, OCLC 0123456789, PMC 0123456789, PMID 0123456789, ID0123456789

And that would be redundant with JSTOR 0123456789

Include other identifiers

Identifiers with current transclusion/ raw url count.

3

This would allow us to have them properly ordered, rather than placed between the doi and the ISBN (current behaviour) or appended at the very end (post-cleanup behaviour) through |id={{arxiv|0123456789}} or similar. I note that they most are found in {{citation/identifier}}, but these somehow aren't used by the main template. It would also probably facilitate bot-maintenance. It might not be possibly to have them all due to technical limitations, but it should at least be possible to include several of them.

Note that these numbers do not include "raw urls" or use through other templates such as {{citation|id=JFM 0123456789}} or {{cite arxiv|eprint=0123456789}}. For example, {{OSTI|NUMBERS}} is transcluded 15 times, but the "raw url" equivalent (http://www.osti.gov/energycitations/product.biblio.jsp?osti_id=NUMBERS) is present 166 times. {{OL|NUMBERS}} is transcluded 7 times, but the "raw url" equivalent (http://openlibrary.org/b/OLNUMBERSM) is present 789 times.

Comments

Opinions? Headbomb {talk / contribs / physics / books} 02:10, 16 February 2011 (UTC)

All of this sounds like a good idea to me. —David Eppstein (talk) 06:39, 16 February 2011 (UTC)

Pubmed seems to be a Pubmed Entrez search term rather than an identifier?

I'd suggest suppressing ISSN, LCCN if provided, if there's a working PMID, PMC, DOI or JSTOR — keep ISSN and OCLC links for the less commonly available journals, and or use a Bot to link the journal field to it's Wikipedia page rather than displaying an ISSN or LCCN.

It would be good to do the same for books (suppress OCLC, LCCN) — if an ISBN is present, but defining a working ISBN seems to be hard for less mainstream publishers. RDBrown (talk) 07:21, 16 February 2011 (UTC)

That's worth investigating, but that should probably be done in another section so we don't have two discussion in the same thread. Headbomb {talk / contribs / physics / books} 07:59, 16 February 2011 (UTC)

Trying to be a bit more on topic, the IDs group a bit by the ref, and it's containing entity.

So for a journal article the PMC gives the free article (links the title if no url), DOI would if you have access, PMID, arXiv and more would link an abstract.

OCLC may link the article or the journal, ISSN and LCCN would link the journal.

— so maybe order IDs by relating to ref, followed by relating to containing publication.

Books seem simpler, though some online books/encyclopedias can have DOIs at the chapter level, a set and volume can each have an ISBN and some series may have an ISSN too. Ref changes have been an education that way. RDBrown (talk) 08:48, 16 February 2011 (UTC)

I like the idea of standardizing the order, though the usual rule of making only edits with visible results still pertains. I'm not sure I buy the whole "clutter" as a problem, but I'll accept that others think so and respect their esthetic preferences. If there is priority to apply (suppressing linkage of some IDs when others are present) we have to make choices based on principles. Do we want to help readers find the exact version cited at the cost of making it more difficult to find some version? If so then suppress OCLC for ISBN, suppress OCLC for ISSN, suppress PMID for DOI, etc. I lean the other way: wp:WORLDWIDE doesn't have to conflict with wp:SAYWHEREYOUGOTIT. I think we should rather help editors find a website or library that has the obscure work, even if it's an earlier or later edition, or even the same edition on a different press, and hence bearing a different ISBN. By all means use the specific source to verify the exact and detailled publication data displayed, but let's link to the more generally (and ideally freely) accessible source. LeadSongDog come howl! 16:10, 16 February 2011 (UTC)

How often are these identifiers used? I'm not convinced we need a specific parameter for identifiers that are only used a few tens of times. I don't see a big problem in the use of |id= and the appropriate mini template for uncommon identifiers. Rjwilmsi 17:50, 16 February 2011 (UTC)

Most of them (like arxiv, ASIN, JFM, OSTI, OL, LCC, LCNN, MR, PubMed, SSRN, UPC, Zbl, ....) are used several hundred if not thousands of times. The "low use" ones would be AFRI, EEEC, NDL, and RFC, but it doesn't really cost anything to include them in the bunch. It's not so much that it's a "problem to use |id=", than it's "since we're overhauling identifiers, might as well go all the way". Headbomb {talk / contribs / physics / books} 19:03, 16 February 2011 (UTC)

In fact there is a problem with using |id= in that it can't really support multiple values for a single citation. Readers and editors have access to various repositories, often non-overlapping. Where one might be able to use {{NDL}} but not {{JFM}}, another might have the reverse situation. Hence we serve them better by identifying the citation in as many ways as practicable. But if we have to use id={{NDL}} then we can't also use id={{JFM}}. LeadSongDog come howl! 02:21, 24 February 2011 (UTC)

It can support multiple values for a single citation; but to do this you need to bear in mind a restriction that applies to all named parameters in all templates: any given named parameter may only be provided once per template - if a named parameter is given more than once, the last one overrides all the others. Thus:

|id={{NDL|21005395}} |id={{JFM|26.0429.03}}

is treated as if it were:

|id={{JFM|26.0429.03}}

Since the content of |id= is not processed in any way by {{citation}} but passed on to {{citation/core}} (through its |ID= parameter), and {{citation/core}} doesn't process |ID= either, merely outputting it as it stands, save for the possible insertion of a comma and space beforehand (it's not even put into the COinS metadata at the end), it follows that |id= is free-form text and therefore may include multiple identifiers. We can therefore use:

|id={{NDL|21005395}}, {{JFM|26.0429.03}}

which covers the requirement of LeadSongDog. --Redrose64 (talk) 11:32, 24 February 2011 (UTC)

True, but these will still won't be well-integrated with the rest of the template (no COinS, seperate treatment dependent on other templates, presented out of sequence, ...). A case can certainly be made against including {{AFRI}} (very very rare), but there's no reason to exclude those with significant present on Wikipedia, such as OSTI and JFM/Zbl. Headbomb {talk / contribs / physics / books} 11:49, 24 February 2011 (UTC)

In some cases it may be that there is a many:one mapping rendering the rare id types unnecessary (except as a transient measure). Major national libraries such as Japan's NDL may have their catalogues entirely mapped to some more universal system such as OCLC. Where the more general system of a union catalogue is available (and ids are confirmed by the specific catalogue) I see no problem in supplanting the latter's use. LeadSongDog come howl! 14:02, 24 February 2011 (UTC)

I have to come think that "no COinS" would be a feature. - J. Johnson (JJ) (talk) 23:45, 4 March 2011 (UTC)

I doubt you'll have consensus for that. Especially for a "COinS, but only for some identifiers". Headbomb {talk / contribs / physics / books} 20:16, 6 March 2011 (UTC)

I hardly expect such a position to find favor amongst the proponents of COinS, who are much enamored of its benefits. No one speaks of the detriments (aside from myself, it seems) because no one (?) sees them as due to COinS. What people see is a citation problem (with the {{citation}} and {{cite xxx}} templates). In particular, a performance problem. This is one of the main reasons people don't use the templates. Perhaps there is a better way to deliver the funtionality of COinS. As is, all the effort of implementing COinS is rather pointless if it results in rejection of the feature it is intended to enhance. If editors were required to use the templates, and were aware of the cause of the performance problems, I suspect there would be a strong consensus to remove COinS. - J. Johnson (JJ) (talk) 21:29, 6 March 2011 (UTC)

Implement jstor tweaks / alphabetization now?

{{edit protected}} Could the jstor tweaks and the alphabetization be implemented? I'd propose a specific change, but I'm not very familiar with how {{citation}} is structured / works. This seems to have consensus, and implementing them now would allow us to focus on the discussion about which identifiers to include. Headbomb {talk / contribs / physics / books} 10:03, 23 February 2011 (UTC)

Anyone? Headbomb {talk / contribs / physics / books} 19:54, 4 March 2011 (UTC)

Alright, I gave a shot at this and made some improvements on the side (all tested). All that needs to be done is to copy {{citation/sandbox}} to {{citation}} and {{citation/core/sandbox}} to {{citation/core}}. What this change is:

Identifiers are alphabetized, with the custom ID placed last.
JSTOR > PMC > URL is now streamlined in {{citation}}. Before, JSTOR > URL was implemented in {{citation}}, and PMC > URL was implemented in {{citation/core}}.
JSTOR / PMC urls are now suppressed in print.
The JSTOR and PMC are now also displayed online.
COinS stuff is untouched

It's all tested, and works like a charm as far as I'm aware. Headbomb {talk / contribs / physics / books} 20:47, 4 March 2011 (UTC)

Done --Closedmouth (talk) 20:04, 6 March 2011 (UTC)

Implementation

{{edit protected}} Alright, based on the above discussion, I've made the tweaks which I believe have consensus (see {{citation/sandbox}}/{{citation/core/sandbox}}/{{citation/identifier/sandbox}}, all tested and which work like a charm as far as I'm aware), as well as other minor cosmetic tweaks (bibcode currently has a colon, which is inconstant with the rest). Namely, these identifiers would be fully supported (meaning alphabetization, COinS, online links, offline no-links):

3

While these would not

3

The first number represent the number of articles which transcludes a specific identifier templates (such as {{ASIN}}), while the second represents the number of articles the raw url associated with the identifier template is found (for {{ASIN}}, this means "http://www.amazon.com/dp/0123456789"). This represents a lower-limit, as these can be found multiple times per article, and url variations exist. The cutoff for inclusion is roughly 150 hits. Specifically the "least used" included identifier is OSTI with 167 uses (JFM/Zbl count as one identifier, since it's two version of the same database/url, totaling 252 combined hits), and the "most used" non-included one is LCC (with 35 hits). Headbomb {talk / contribs / physics / books} 07:42, 7 March 2011 (UTC)

Done Killiondude (talk) 09:02, 7 March 2011 (UTC)

Great to see this done...any news on a Google book version?Moxy (talk) 02:45, 8 March 2011 (UTC)

A what? Headbomb {talk / contribs / physics / books} 03:34, 8 March 2011 (UTC)

Can we get arXiv as a synonym for arxiv? I feel that this will be used often. It should be as simple as replacing {{{arxiv|foo}}} with {{{arxiv|{{{arXiv|foo}}}}}}. CRGreathouse (t | c) 16:02, 16 March 2011 (UTC)

Makes sense to me. Headbomb {talk / contribs / physics / books} 16:31, 16 March 2011 (UTC)

Part and section

As far as I can tell there is neither a part of a section parameter. Three questions

Has this been discussed before?
As there is a volume= parameter why no part= or section= parameters?
What is the work around for placing these three books all cited in the article Seige into templates which are usable with {{harv}}?
- Needham, Joseph (1986). Science and Civilization in China: Volume 4, Part 2. Taipei: Caves Books Ltd.
- Needham, Joseph (1986). Science and Civilization in China: Volume 5, Part 6. Taiepi: Caves Books Ltd.
- Needham, Joseph (1986). Science and Civilization in China: Volume 5, Part 7. Taipei: Caves Books Ltd.

-- PBS (talk) 14:35, 14 March 2011 (UTC)

The |volume= parameter doesn't actually put the word "volume" in there, so you can use it for parts/sections, but it looks kinda weird:

*{{citation |last=Needham |first=Joseph |year=1986 |title=Science and Civilization in China |volume=Volume 4, Part 2 |location=Taipei |publisher=Caves Books Ltd }}

Needham, Joseph (1986), Science and Civilization in China, vol. Volume 4, Part 2, Taipei: Caves Books Ltd {{citation}}: |volume= has extra text (help)

so what I (and others) do is to put it all in the title:

*{{citation |last=Needham |first=Joseph |year=1986 |title=Science and Civilization in China: Volume 4, Part 2 |location=Taipei |publisher=Caves Books Ltd }}

Needham, Joseph (1986), Science and Civilization in China: Volume 4, Part 2, Taipei: Caves Books Ltd

There are two common techniques for enabling Harvard linking where the author(s)/year taken together are not unique. One is to suffix the year with a small letter:

*{{citation |last=Needham |first=Joseph |year=1986a |title=Science and Civilization in China: Volume 4, Part 2 |location=Taipei |publisher=Caves Books Ltd }}

Needham, Joseph (1986a), Science and Civilization in China: Volume 4, Part 2, Taipei: Caves Books Ltd

so that you can use {{harv|Needham|1986a|p=123}} (Needham 1986a, p. 123). The other technique is to customise the ref link using {{harvid}}:

*{{citation |last=Needham |first=Joseph |year=1986 |title=Science and Civilization in China: Volume 4, Part 2 |location=Taipei |publisher=Caves Books Ltd |ref={{harvid|Needham|1986 vol 4 pt 2}} }}

Needham, Joseph (1986), Science and Civilization in China: Volume 4, Part 2, Taipei: Caves Books Ltd

so that you can use {{harv|Needham|1986 vol 4 pt 2|p=123}} (Needham & 1986 vol 4 pt 2, p. 123) --Redrose64 (talk) 16:01, 14 March 2011 (UTC)

Didn't know about {{harvid}} so thanks for the information.

regarding the citation template, if all one is going to do is wrap volume, part and section up into the title, then why bother to have a separate volume= parameter as one could always wrap the volume into the title? AFAICT, the whole point about separating out parameters is so that we can have a uniformed look. suggesting that "Volume 4, Part 2" is placed into the title seems to me a retrograde step, as it removes the use of one parameter and what is added to the title becomes a matter of style choice for the individual editor. -- PBS (talk) 19:27, 14 March 2011 (UTC)

I think volume should not be wrapped into title. But if some publisher wanted to do, say, Volume 4a, or Volume 4xyz, etc., I think there is no problem. And "Volume 4 Part 2" could be considered the same kind of case, differing only in having included spaces. Of course it looks weird, because it is weird, but the onus for that is on the publisher. And we shouldn't sweat trying to make it look prettier than it really is. - J. Johnson (JJ) (talk) 20:07, 14 March 2011 (UTC)

PBS and J. Johnson are correct that the title should not be mangled to include the volume, but you have also confused the series title with the volume title. If necessary, I would use |volume=4|at=Part 2, and I'd also add the OCLC 489880705 so readers can find the work in a library. Hence

{{citation |last=Needham |first=Joseph |year=1986 |series=Science and Civilization in China |title=Physics and physical technology |volume=Volume 4 |at=Part 2: Mechanical engineering |publisher=Cambridge University Press |oclc=489880705 }} which renders:
Needham, Joseph (1986), Physics and physical technology, Science and Civilization in China, vol. Volume 4, Cambridge University Press, Part 2: Mechanical engineering, OCLC 489880705 {{citation}}: |volume= has extra text (help)

Similarly for OCLC 38091229 and OCLC 417656050

{{citation |last=Needham |first=Joseph |year=1994 |series=Science and Civilization in China |title=Chemistry and chemical technology |volume=Volume 5 |at=Part 6: Military technology : missiles and sieges |publisher=Cambridge University Press |oclc=38091229}} which renders:
Needham, Joseph (1994), Chemistry and chemical technology, Science and Civilization in China, vol. Volume 5, Cambridge University Press, Part 6: Military technology : missiles and sieges, OCLC 38091229 {{citation}}: |volume= has extra text (help)

and

{{citation |last=Needham |first=Joseph |year=1986 |series=Science and Civilization in China |title=Chemistry and chemical technology |volume=Volume 5 |at=Part 7: Military technology: The gunpowder epic |publisher=Cambridge University Press |oclc=417656050}} which renders
Needham, Joseph (1986), Chemistry and chemical technology, Science and Civilization in China, vol. Volume 5, Cambridge University Press, Part 7: Military technology: The gunpowder epic, OCLC 417656050 {{citation}}: |volume= has extra text (help)

— Preceding unsigned comment added by LeadSongDog (talk • contribs) 20:44, 14 March 2011

(edit conflict) {{citation}} is a general-purpose one-template-fits-all tool, hence it has many parameters which will be irrelevant to the particular information source which it is being applied to; and it has other parameters which are relevant, but which might not quite format the information as you would wish. The |volume= parameter is there primarily for journal citations, and it's not mandatory to use it in other publications which also have a volume number. Whether or not |volume= should be used for books has been discussed before both here, at template talk:citation/core and at template talk:cite book: I don't recall a firm consensus being reached. Look through these three talk pages and their archives (particularly for cite book), and you'll find many discussions headed something like volume number should not be in bold. --Redrose64 (talk) 20:53, 14 March 2011 (UTC)

Of course we need a parameter for volume for books. The Dictionary of National Biography is an example without a volume it would be difficult to auto-date the publication. I think that as we have edition information and volume information it would be advantageous for similar reasons to have a part or section parameter as well. -- PBS (talk) 19:17, 15 March 2011 (UTC)

I am somewhat concerned that adding features (parameters) rather ad hoc for specific cases tends to increase both the complexity of the template, and (importantly!) the existing perception of excessive complexity. I don't know what subtleties of bibliographic practice affect "series" and "part" in different contexts, and am concerned that making them (or any other parameter) behave in specific ways over a range of contexts would be contortuous. I wonder if we could have a parameter "xxx" such that we could say (e.g.): it is exactly like "vol" except not bolded, and follows "vol" where both are present; it is useful for names/numbers of book volumes, or series titles. - J. Johnson (JJ) (talk) 21:51, 15 March 2011 (UTC)

I suppose this is yet another example of needing to follow and build on the best practices already being used elsewhere. We should ask ourselves which fields are supported in other bibliliographic citation tools. LeadSongDog come howl! 22:32, 15 March 2011 (UTC)

Preprints

Could a preprint field be added to this template? The result should look as follows:

Knuth, Donald E. (December 2003). "Robert W Floyd, In Memoriam". ACM SIGACT News. 34 (4): 3–13. {{cite journal}}: Invalid |ref=harv (help) (preprint)

Preprints have the advantage that they are usually available for free to anyone, but might differ from the (paywalled) official version that you want to link from using the url field and should be clearly marked as such. —Ruud 12:51, 23 February 2011 (UTC)

This seems analogous to the notation "[abstract]", except for the explicit url. I usually let the link to the official (typically non-free) version by means of the doi, and use the url= parameter to link to any free version I can find. Which can differ; I hadn't thought it necessary to point that out. There can also be postprints, with corrections and supplementation; should that be another field? - J. Johnson (JJ) (talk) 00:16, 24 February 2011 (UTC)

For the record, I support this request. It's no big deal to implement and saves typing. The alternative is to use the postscript parameter, but everyone is going to have their own variation of that, and this could be an issue in FAs etc. I had raised this myself at Template_talk:Cite_journal#Preprint.3F. Tijfo098 (talk) 20:21, 23 March 2011 (UTC)

Some clarification would be needed to meet wp:SAYWHEREYOUGOTIT. Which version is cited, and which is the "also available at"? LeadSongDog come howl! 21:08, 23 March 2011 (UTC)

I don't think that's an issue. Preprints do not differ much from the officially published version. They certainly differ less than a science paper differs from the journalistic take that is added with |laysummary= Tijfo098 (talk) 21:50, 23 March 2011 (UTC)

Why not just use |id={{preprint|url=http://www-cs-faculty.stanford.edu/~uno/papers/floyd.ps.gz |format=PostScript}} to give Knuth, Donald E. (December 2003). "Robert W Floyd, In Memoriam". ACM SIGACT News. 34 (4): 3–13. {{preprint}}. {{cite journal}}: Invalid |ref=harv (help)? Seems like the simplest way to do it, no? Headbomb {talk / contribs / physics / books} 00:14, 24 March 2011 (UTC)

There are many parameters that you could subvert to put arbitrary text in them; |id= is one an so is |postscript=. My point is that neither of those is very intuitive or user friendly for what I see as a common usage pattern. [4] Tijfo098 (talk) 04:02, 29 March 2011 (UTC)

Columns

Is it possible to include a column entry? I often cite from old newspapers, which basically are broadsheet-sized pieces of paper filled edge-to-edge with small text. Columns are often provided in the source's citation (see The Times archive), it would be very useful to be able to include columns in citations here. Parrot of Doom 08:33, 20 May 2011 (UTC)

In such cases I use the |at= parameter instead of the |page= or |pages= parameter. For example, |at=p. 12, col. 3 --Redrose64 (talk) 13:21, 20 May 2011 (UTC)

Zbl must be "zbl"

The updated template is great, so thanks again. Mathematical Reviews works fine. However, the German Zbl functioned via the id subterfuge, not the "Zbl=" commands. Thanks, Kiefer.Wolfowitz (Discussion) 20:52, 27 March 2011 (UTC)

ID= (Works):

Illés, Tibor; Szirmai, Ákos; Terlaky, Tamás (1999). "The finite criss-cross method for hyperbolic programming". European Journal of Operational Research. 114 (1): 198–214. doi:10.1016/S0377-2217(98)00049-6. ISSN 0377-2217. Zbl 0953.90055. PDF preprint. {{cite journal}}: Cite has empty unknown parameter: |1= (help); Invalid |ref=harv (help)

Zbl=Blah (Buggy):

Illés, Tibor; Szirmai, Ákos; Terlaky, Tamás (1999). "The finite criss-cross method for hyperbolic programming". European Journal of Operational Research. 114 (1): 198–214. doi:10.1016/S0377-2217(98)00049-6. ISSN 0377-2217. PDF preprint. {{cite journal}}: Cite has empty unknown parameters: |1= and |2= (help); Invalid |ref=harv (help); Unknown parameter |Zbl= ignored (|zbl= suggested) (help)

You have to use a lowercase zbl, mr, doi, jstor, etc... Headbomb {talk / contribs / physics / books} 04:19, 29 March 2011 (UTC)

zbl= (WORKS!):

Illés, Tibor; Szirmai, Ákos; Terlaky, Tamás (1999). "The finite criss-cross method for hyperbolic programming". European Journal of Operational Research. 114 (1): 198–214. doi:10.1016/S0377-2217(98)00049-6. ISSN 0377-2217. Zbl 0953.90055. PDF preprint. {{cite journal}}: Cite has empty unknown parameter: |1= (help); Invalid |ref=harv (help)

Thanks for the correction. The lower case "zbl= " works fine. Kiefer.Wolfowitz 13:12, 21 May 2011 (UTC)

BTW, you shouldn't link to science direct. The article isn't freely available there, and the link is redundant with the DOI. Also the preprint is not in PDF, but in PostScript format. And the ISSN should be omitted as well, it's got nothing to do with the article. AKA aim for

Illés, Tibor; Szirmai, Ákos; Terlaky, Tamás (1999). "The finite criss-cross method for hyperbolic programming". European Journal of Operational Research. 114 (1): 198–214. doi:10.1016/S0377-2217(98)00049-6. Zbl 0953.90055. .ps preprint. {{cite journal}}: Invalid |ref=harv (help)

or something similar Headbomb {talk / contribs / physics / books} 13:38, 21 May 2011 (UTC)

I should provide ISSN's only for rare (or often out-of-print) journals, which this isn't, as you note. Kiefer.Wolfowitz 13:43, 21 May 2011 (UTC)

I understand your complaint that url= is redundant with doi= here, but don't understand the point of your stating that the article is not free. Kiefer.Wolfowitz 13:46, 21 May 2011 (UTC)

RfC to add dead url parameter for citations

A relevant RfC is in progress at Wikipedia:Requests for comment/Dead url parameter for citations. Your comments are welcome, thanks! — HELLKNOWZ ▎TALK 11:00, 21 May 2011 (UTC)

Explain mr, zbl, jfm

The template has mr, zbl, and jfm as the last three parameters, and there's discussion on this page about capitalization of zbl (#ZBL → Zbl, #Zbl must be "zbl"), but what the heck are they? The example has nothing about them at all. --Thnidu (talk) 15:58, 28 May 2011 (UTC)

Parameters for particular types of identification numbers, just as |isbn=, |issn= and |doi= are. --Redrose64 (talk) 17:08, 28 May 2011 (UTC)

MR is Mathematical Reviews, ZBL is Zentralblatt MATH and JFM is Jahrbuch über die Fortschritte der Mathematik which redirects to Zentralblatt MATH. ---— Gadget850 (Ed) ^talk 17:21, 28 May 2011 (UTC)

trans_title

There is a parameter trans_title listed in the section with the rather long title Citing edited books, or parts of edited books, including encyclopedias and encyclopedia articles which doesn't seem to work. The format including an underscore is different to all the others and I'm wondering if it really exists. If it does, what's its name?

Also is there any benefit in the long title? Would it be better simply as "Citing parts of books, including encyclopedia articles"? Chris55 (talk) 18:00, 30 May 2011 (UTC)

You can tell very easily if a parameter exists or not by looking at the template content. Go to the template page (Template:Citation), click on the view source tab, and search for three opening braces followed by the parameter name, ie search for {{{trans_title. It's definitely there, so is definitely valid.

Concerning why it doesn't work: you don't describe the test data which you have used. It does work, but only in certain situations. --Redrose64 (talk) 19:13, 30 May 2011 (UTC)

As far as I can see from the template page it is only defined for patents. Chris55 (talk) 20:48, 30 May 2011 (UTC)

Organisation of the page

Does it help having the parameters repeated 4 times? They all seem to have grown to include most of the parameters but there are significant discrepancies between them and will continue to be. I've tried to bring some of the advantages of the "periodicals" to the "books" section and it would make maintenance easier to simply have 1 full parameter section and 4 examples sections.

To make it easier for newbies to get started, would it help to also have a simple section at the beginning with the most commonly used parameters. e.g.

{{Citation \|last= \|first= \|editor= \|year= \|title= \|publication-place= \|publisher= \|page= \|isbn= \|url= \|accessdate= }}	last: The author's surname or last name. (Don't use with author.) first: The author's first or given name(s). editor: Name of editor, placed after date with `,ed.` year: Year of authorship or publication. Mandatory for use with links from Template:Harvard citation. Don't use with date. title: Title of the work. Mandatory for web references. edition: Number or name of the edition, if not the first; for example: edition=2nd. publication-place (or place): The city of publication. If more than one town/city is listed on the title page, give the first one or the location of the publisher's head office. publisher: The name of the publisher. Omit terms such as Publishers, Co., Inc., Ltd., etc., but retain the words Books or Press. page: For use when one page is cited. Adds "p." before the page number. Do not use with pages. isbn: Use this parameter if a book which has an ISBN; see Wikipedia:ISBN for guidance on format. url: An url of an online location where the item can be found. access-date: Date when the url was accessed.

I've deliberately left in some references to other ways of doing it so that people realise there are alternatives. If it needs to be even simpler there are one or two that could be dropped. Chris55 (talk) 19:02, 30 May 2011 (UTC)

They are deliberately separate, because the rendered appearance is quite different. Except for capitalisation and punctuation, one is supposed to give a result similar to {{cite journal}}, another similar to {{cite book}}. --Redrose64 (talk) 19:16, 30 May 2011 (UTC)

I'm aware the rendered appearance is different, at least for books and periodicals, but I don't see that the documentation is necessarily helped by separating them. The descriptions are a mess: neither summary or full list is complete in most cases; the ordering of the two sides is mixed up; it's not clear at all which parameters can be used for what purpose. The examples will show the different output and people can be pointed by a suitable introduction. Chris55 (talk) 20:46, 30 May 2011 (UTC)

I should make it plain that I don't include patents in the joint parameter list as there's little commonality there. But if there are no other objections I'll try and show how they could all fit together. There are very few differences: parts have a contribution attribute and periodicals don't need to include editors and a few similar fields. Apart from that, most of the differences are errors in the documentation. Chris55 (talk) 09:08, 31 May 2011 (UTC)

PMC URL generation discussion

Editors here may be interested in Template_talk:Cite_journal#Recent_edit_to_remove_URL_generation_from_PMC_parameter. Thanks Rjwilmsi 14:42, 12 June 2011 (UTC)

Italicising web page names

Is there any logic to italicising the title of a web page rather than enclosing it in quotation marks, which would seem to be the more natural option?

{{citation |title=Great Moreton Hall, Moreton Cum Alcumlow |url=http://www.britishlistedbuildings.co.uk/en-56513-great-moreton-hall-moreton-cum-alcumlow |publisher=British Listed Buildings |accessdate=10 June 2011}}
Great Moreton Hall, Moreton Cum Alcumlow, British Listed Buildings, retrieved 10 June 2011 {{citation}}: Check date values in: |accessdate= (help)

It's possible to force what seems to be the correct formatting like so:

{{citation |title=''"Great Moreton Hall, Moreton Cum Alcumlow"'' |url=http://www.britishlistedbuildings.co.uk/en-56513-great-moreton-hall-moreton-cum-alcumlow |publisher=British Listed Buildings |accessdate=10 June 2011}}
"Great Moreton Hall, Moreton Cum Alcumlow", British Listed Buildings, retrieved 10 June 2011 {{citation}}: Check date values in: |accessdate= (help),

so it's not critical (although I've just noticed that leaves a trailing comma), but it's inconvenient and an obvious incompatibity with {{cite web}}.

{{cite web |title=Great Moreton Hall, Moreton Cum Alcumlow |url=http://www.britishlistedbuildings.co.uk/en-56513-great-moreton-hall-moreton-cum-alcumlow |publisher=British Listed Buildings |accessdate=10 June 2011}}
"Great Moreton Hall, Moreton Cum Alcumlow". British Listed Buildings. Retrieved 10 June 2011. {{cite web}}: Check date values in: |accessdate= (help)

Malleus Fatuorum 15:22, 12 June 2011 (UTC)

Entire works (books, journal titles, etc) are italicized; subsidiary works within them are in quotation marks. In the cases you list, the web page is taken to be the entire work. If you want to show it as a part of a larger web site, you could use contribution= instead of title=:

{{citation |contribution=Great Moreton Hall, Moreton Cum Alcumlow |url=http://www.britishlistedbuildings.co.uk/en-56513-great-moreton-hall-moreton-cum-alcumlow |publisher=British Listed Buildings |accessdate=10 June 2011}}

"Great Moreton Hall, Moreton Cum Alcumlow", British Listed Buildings http://www.britishlistedbuildings.co.uk/en-56513-great-moreton-hall-moreton-cum-alcumlow, retrieved 10 June 2011 {{citation}}: Check date values in: |accessdate= (help); Missing or empty |title= (help) —David Eppstein (talk) 15:51, 12 June 2011 (UTC)

Ah, OK. I don't really see the logic in considering a single web page to be an entire work, but your workaround fixes the problem I had, so thanks for that. Malleus Fatuorum 16:31, 12 June 2011 (UTC)

why do we wikilink "doi" next to the actual doi external link?

It looks like crap and is creating a gazillion extra wikilinks. Just link the doi itself. The whole citations are turning into a mash of spread together blue with this sort of thing (people wikilinking journals and publishers even, also). Please stop. Leave the link for high value usage like the one that actually goes to the article!TCO (talk) 09:56, 15 June 2011 (UTC)

Per WP:MULTI, please link to your original post at Template talk:Cite journal#why is "doi" wikilinked? rather than restating your point in a different form. --Redrose64 (talk) 13:02, 15 June 2011 (UTC)

Sorry, I wasn't meaning to spam...was kind of flailing around last night. :( TCO (talk) 13:03, 15 June 2011 (UTC)

Translated titles

I know for Cite web, there is a string for titles of works that have been translated into English (using the string |trans_title =). Is there any way this could be added to the template (so it appears after the original title)? User:Zscout370 ^{(Return Fire)} 21:13, 27 July 2011 (UTC)

I thought it had been: but looking more carefully, it's been put in the wrong place - it's not been put within {{Citation/core}} but within {{Citation/patent}}, which does not recognise these parameters. Here's the fix: {{editprotected}} Within Template:Citation, please remove from the {{Citation/patent}} transclusion the following two lines:

|TransTitle={{{trans_chapter|}}}
|TransItalic={{{trans_title|}}}

and insert them within the {{citation/core}} transclusion, preferably immediately after the

|Title={{{title|}}}

for consistency with {{cite book}}. At that point there is the following line:

|Trans_title ={{{trans_title|}}}

which was added with this edit - this line should also be removed, as it does nothing. --Redrose64 (talk) 22:18, 27 July 2011 (UTC)

It is needed in the general citation area. I can make the fixes. User:Zscout370 ^{(Return Fire)} 22:55, 27 July 2011 (UTC)

[1] IBM & SA22-7085-1, p. 12345. A Møøse once bit my sister ...

[2] IBM & SA22-7085-1, pp. 1234567–12345678. No realli! She was Karving her initials on the møøse ...

[1]

[2]

@@ Line 1,454: / Line 1,454: @@
 *{{NDL|0123456789}} x14 / 1
 *<nowiki>{{OL|0123456789}}</nowiki> x7 / 745
-*{{OSTI|0123456789}} x15 / 152
+*<nowiki>{{OSTI|0123456789}}</nowiki> x15 / 152
 *<s>{{PubMed|0123456789}}</s>  (redundant with {{para|pmid}}) x24
 *{{Rfcurl|0123456789}} x4 / 245
@@ Line 1,525: / Line 1,525: @@
 *{{MR|0123456789}} x1182 / 36
 *<nowiki>{{OL|0123456789}}</nowiki> x7 / 745
-*{{OSTI|0123456789}} x15 / 152
+*<nowiki>{{OSTI|0123456789}}</nowiki> x15 / 152
 *{{Rfcurl|0123456789}} x4 / 245
 *{{SSRN|0123456789}} x21 / 524