Template talk:Cite book/Archive 1

From Wikipedia, the free encyclopedia

 This is an archive of past discussions. 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.
Archive 1 Archive 2 →

Contents

Backwards compatibility

We need backwards compatibility with book reference. Please do not make things like this before discussing it. I have thus reverted. --Ligulem 15:27, 16 February 2006 (UTC)

I was thinking as you converted you could also convert the field names. But I suppose that makes settting up AWB quite difficult. ··gracefool |☺ 20:33, 16 February 2006 (UTC)
We can convert parameter names while upgrading. That's not a problem at all. BTW you, anybody can do that too with WP:AWB (it's not such a big thing, just start slowly and check each diff carefully. I also check each diff - oh my eyes are hurting already after doing that cite journal migrate thing :-). It's just a little bit regex fiddleing (I will provide the AWB settings once I've started, so we can share it and learn from each other). What I argue is, do we really need to add more parameters for additional authors (which are currently put together in the coauthors param)? Is it really needed to abandon the nice short "first" "last" parameters? If we do not need to handle the additional authors separately, then there is no need to abandon "first" "last". Handling additional authors separately adds considerable complexity to the template. What do we gain by doing that? On the other hand, you are correct that this would be the right moment to do such a step. What I fear is that we pile up params here and they will not be used. Params once added are very hard to remove later, even when used seldom. --Ligulem 22:21, 16 February 2006 (UTC)
See the new documentation. All first and last fields are eliminated, as they are unneccessary. Using author2, author3 etc. means the template controls all the punctuation, ie. the commas between each author, and the ampersand between the last authors. Note I've also eliminated the authorlink field, as well as any use of {{wikilink}} - these are also unneccessary (see Template talk:wikilink). I've been working on keeping backwards compatibility with deprecated fields, but do you think I can get rid of them now? ··gracefool |☺ 22:35, 16 February 2006 (UTC)
Hmmm. I see you would like to deprecate (or even remove) first last parameters. When I was a book ref newbie, I once proposed that on the old book reference but SEWilco was against that and COGDEN also voiced for first last. SEWilco then added the coauthors parameter. See Template talk:Book reference/archive01#Eliminating variant First, Last. Question is: shall we revisit that decision taken then? The reasoning of SEWilco then for first last was a bit vague. But he might have had other reasons he didn't say explicitly. Does anybody know what the real striking argument(s) were that lead to that first, last thing? I think I might be a bit blinded by past consensus finding sessions. Gracefool seems to come to the same conclusion I once had. But I don't know what opinion I have now (!). I need to ruminate a bit more on this... --Ligulem 23:22, 16 February 2006 (UTC)
A possible argument for using first last I see is that on book references, the first author is written as
last, first
whereas all following authors are written as
first last, first last, ...
So the first author is treated somewhat special (See the multi author examples on {{book reference}}). --Ligulem 23:36, 16 February 2006 (UTC)
  • Yes, however you could just specify the order in the documentation (as in the devel cite book). The only benefit I can think of (after reading that archive page) is for controlling formatting, and preventing user errors, which is good. However, to do it properly you also need an authortitle field, a authorseniority field, etc. That all increases server load, and decreases ease-of-use and developer maintainability. COGDEN also pointed out that not all cultures put the surname last normally, so in any case the field names need to change (eg. to surname and given). But I think it's best to remove them entirely, for those reasons. ··gracefool |☺ 23:58, 16 February 2006 (UTC)
  • A benefit of first last for the #1 author is also that books are usually "addressed" by the last name of that author and the year of the book. So if book references might once get migrated to a database, we already have that info. We already have quite a number of book ref calls that actually use first, last. If we go back, this is a destruction of information on the calls. This is sometimes a good hint. BTW parameter first recieves everything of the name of the author that is not "last". So for example for "Robert Cecil Martin", first is "Robert Cecil" and last is "Martin". If he is the primary author of a book, then it is "Martin, Robert Cecil". Coauthors follow with reversed name parts. --Ligulem 00:39, 17 February 2006 (UTC)
True that. I'll reinstate first and last, and deprecate author.
As for author1 to author5 - they are optional, since there's no way to stop people using coauthors instead. So it's not a problem for people who don't like them. Their code isn't complex (it's just repeated 4 times), and although you're right that they're not needed, they are nice and make for more consistency. ··gracefool |☺ 09:11, 19 February 2006 (UTC)
I agree with the deprecation of author in favor for first last. Per the author1 to author5 I'm still rather on the opposing end. It does make the template more complicate and it is not a complete solution (5 is just an arbitrary number). I'm also still quite a bit paranoid about the stability of qif (see User talk:Brion VIBBER#The future of qif). He doesn't like qif (he was even more explicit in PM) and he's the CTO of Wikipedia (and besides that a very hard working damned knowledgeable guy. After all he keeps our buts running here). It's still a bit a desaster, I know. Qif is the least ugliest thing we have for now. But if we can limit it's use, we should do so (but not at any cost of course, so to hell with the policy tag on WP:AUM). There are also some people that are strong opponents to using book ref (cite book) in articles because they feel it is needlessly complex (which I do not think so). The lesser we go over board with qif the better (or the lesser ugly?). If qif breaks before we have built in support I might throw in weeble code or hiddenStructure hack here as a stop gap measure (as we had to do here when pressure exploded from WP:AUM). The more complicate this here is, the smaller the chances are we would be able to do so. But at least I'm not comfortable with that qif too and I would love to have this darned beloved thing finally burned forever, but not until we have a proper built-in conditional. So I would petition to you to not go the road of author1 to author5 for now. But If you really really really insist on this, I would bite on my teeth and look away when you insert these ugly author1 to author5 in order to keep you on board. --Ligulem 10:14, 19 February 2006 (UTC)
Okay, I'll go with you. There's no point trying to be perfect, since it's just a stopgap measure anyway. I'll remove all author fields. That will leave us with a template which is nicer than {{book reference}} but uses less server resources. ··gracefool |☺ 12:41, 19 February 2006 (UTC)
Ok. Thanks for your patience. At least you got 4 new params in: link, accessdate, acessyear and accessmonth (these didn't exist on the old book ref). --Ligulem 13:03, 19 February 2006 (UTC)

Gracefool, on User:Gracefool/cite book you went up to 5 authors. What do you do if you have a book with more than 5 authors? I've seen monstrous author lists on cite journal calls (See Sleeping sickness, reported by Phil). --Ligulem 23:29, 16 February 2006 (UTC)

I went up to 6 authors, but wow. Ok, I've reinstated coauthors for cases where there are more than six authors. I expect that is very rare — do you know of any others?
Yeah. Sadly this is quite often on the journals. For books this happens not that often. But rare cases do not free us from tackling them as well. --Ligulem 00:27, 17 February 2006 (UTC)

Ref

In case you didn't notice, I removed the undocumented ref field. It became the "id" HTML attribute for linking with footnote labels. Being undocumented, I'm pretty sure it has been used very little, and it is unnecessary when using the standard {{note}} or {{note label}} templates. I think this should be removed from all cite templates (see Template talk:Web reference/dev). ··gracefool |☺ 20:51, 16 February 2006 (UTC)

Good idea. Thanks for pointing this out. The ref stuff seems to be a bit relic. I haven't seen where that's used. Someone else? User:SEWilco once said something about that on book reference (BTW, I have not heard anything from him recently, he was once very active on the citation templates). I agree to remove it and try going without. --Ligulem 21:56, 16 February 2006 (UTC)
It's probably only used in {{book reference}}. We'll check when we do the big merge to {{citation}} =) ··gracefool |☺ 22:35, 16 February 2006 (UTC)
Funny you should mention that parameter, I used it just this week for something. Anybody want to look through my recent contributions to find out why? (that's a joke BTW) Given that we are aiming to wrap these things in <ref> tags, etc, do we actually need to have anchors in them? —Phil | Talk 12:59, 17 February 2006 (UTC)
You mean the {{ref}} template? ··gracefool |☺ 09:11, 19 February 2006 (UTC)
No. I think Phil talks about Ævar's citation tags. For an example how that's used in a top shot article see AIDS. --Ligulem 09:36, 19 February 2006 (UTC)

I have included links to these anchors in a number of articles, both manually and using other templates for in-text citations (don't recall which template); I just found this discussion because I noticed a broken Harvard-style reference in the article F-34 tank gun. I don't understand the "relic" argument. This is a feature that was in use, and removing it breaks an indeterminate, but probably substantial number of Wikipedia articles. Gracefool, please restore this functionality. —Michael Z. 2006-05-05 18:31 Z

I've tried to restore that functionality. I hope I got it right. Apologies for having removed that. Seems like we were a bit too bold :P. Please check if you see any errors. I think F-34 tank gun works now. At least we now know an article where that's actually used! --Ligulem 20:59, 5 May 2006 (UTC)
Thanks. I know I've used it elsewhere, but there has since been a spate of changes in response to the controversy over conditionals in templates, and to the introduction of cite.php, so I have no idea how many places it still exists. Anyway, although removing this functionality is not fatal, it's still best to avoid even potential breakage. That said, I wish we had designed this behaviour so that we weren't stuck with it this way, better documented it and written more stringent guidelines on its use. Something to consider in future templates and conventions. Cheers. —Michael Z. 2006-05-06 02:49 Z

Migrating book reference to cite book

I've set up settings for WP:AWB to do the migration from book reference to cite book at User:Adrian Buehlmann/AWB/migrate to cite book. Use at your own risk. You might want to wait a bit using these settings until they have been used a bit. This is very fresh meat. --Ligulem 14:19, 17 February 2006 (UTC)

Hiding author on own page

Would it be greatly disturbing if I were to investigate the possibility of hiding the author's name when on their article? I'm thinking of something like if authorlink = {{PAGENAME}} then hide author. Thoughts? —Phil | Talk 16:25, 17 February 2006 (UTC)

If you do this in a sandbox: no :-). BTW, do we get an equivalent for {{booleq|A|B}} when having built-in conditionals? I assume you need an equality function for this. Please note that as soon as Brion breaks qif (don't ask me why he would), there is an incredible chance that booleq breaks. Please also note that I have no clue how to mimick booleq by using weeble code or hiddenStructure hack. So if shit should happen (qif breaks before we have built in conditionals), we will be lost on cite book if you use booleq. --Ligulem 16:50, 17 February 2006 (UTC)
Can't qif do that? As for qif breaking, when it breaks it shouldn't be hard to port templates to the new conditional. ··gracefool |☺ 09:11, 19 February 2006 (UTC)
I think no. Key point is the string equality test (which depends on the default paramater function and a level of meta template, see {{booleq}}). {{qif}} can't do that (and this holds true until you prove me wrong. Burdon of proof on your side :-). Besides, there is a paranoid chance that qif breaks before we have built-in conditionals (but I cannot see why anyone would be so mean to do that). And some aspects of qif (conditional inclusion) will break near certainly (but we do not need conditional inclusion as we have seen here, provided cite book is kept sufficiently simple). The chance is very high that we get at least a built in conditional that provides the equivalent of Wikipedia:hiddenStructure (suppressing of text, dependant on the existence of a parameter, but without the accessability drawback)--Ligulem 10:30, 19 February 2006 (UTC)

Related pages need updating

If this template replaces {{book reference}}, someone should update WP:CITET. Alan Pascoe 13:46, 19 February 2006 (UTC)

Done. --Ligulem 17:04, 19 February 2006 (UTC)

Usage section be more useful in template itself within noinclude block?

User:JonRoma asked in the edit summary of this edit: "shouldn't Usage section be more useful in template itself within 'noinclude block?"

I would propose to not do this because this here is intended as a rather high use template. The old book reference has more than 5'000 entries on the what links here. Adding the usage on the template page itself leads to more edits on the template. Edits should be done one the template only when really needed. At the moment it's not that bad, but once we have converted all inclusions of book reference to cite book, this will be more important. --Ligulem 22:23, 19 February 2006 (UTC)

Original publication date

Copied from Wikipedia_talk:Template_messages/Sources_of_articles/Generic_citations:

The format for citing books that had an original publication date different than the edition cited isn't supported. Wikipedia:Cite sources#Complete citations in a .22References.22 section suggests this:
Marx, Karl [1867] (1967) Capital: A Critique of Political Economy Vol. I. Edited by Frederick Engels. New York: International Publishers.
If you plugged this into the template it'd look like this:
Marx, Karl ([1867] 1967). Frederick Engels Capital: A Critique of Political Economy Vol. I. New York: International Publishers.
or this.
Marx, Karl, [1867] ([1867] 1967). Frederick Engels Capital: A Critique of Political Economy Vol. I. New York: International Publishers.
Neither is particularly appealing, especially with "authorlinks".

The reason I, and the other editor bring this up is that for Baha'i literature and other Baha'i articles, the original dates of the books are much older than the current publication of the books, and we would like to show when the book was originally authored. So is there a possibility of incorporating an original publication date in the template? -- Jeff3000 23:03, 19 February 2006 (UTC)

Sounds reasonable to look into this further. Oh well, smells like (yet) another parameter (origdate? origyear?). Any volunteer to hack up a proposal template code under her or his user space? Best with inclusion examples (I'm currently not enough bored for doing things like this ;-). --Ligulem 23:59, 19 February 2006 (UTC)
  • Speak of the devil! This is one thing that has bothered me about the citation templates. Often I need to cite a short story that has been republished in a new edition. When the publication year appears beside the author's name, it looks as if the original author was around to see republication of his or her work, which can be especially weird when the author is long-dead! Usually I do a work-around by putting the original publication date in brackets after the "first=name" parameter. However, this is a very awkward solution at best. I would definitely welcome an additional parameter to address this problem.
    -,-~R'lyehRising~-,- 01:03, 20 February 2006 (UTC)
    • I've added origdate, origyear and origmonth for this use. ··gracefool |☺ 03:35, 20 February 2006 (UTC)
Thanks for looking into this and adding the new feature. -- Jeff3000 17:22, 20 February 2006 (UTC)
Is it actually useful to have the dates linked? Seems unnecessary to me. —Serein 17:49, 20 February 2006 (UTC)

Editor and Translator

Is there a reason why the "editor" field does not append "Ed." to the name? Also there should be a field for a translator. -- Jeff3000 23:03, 19 February 2006 (UTC)

First question: I don't know (I haven't thought about that deeply enough yet). But we have the problem that the calls now do add this, as mandated by the doumentation. So we cannot add "Ed." on the template, as we would have this twice then.
Second question: There is a parameter "others" (see the doc) that may be used for this. We could add a translator parameter, but I would be happy if this is not needed. I generally prefer not to pile up parameters. We already have a lot. But if there is a very good reason to add a parameter, it should be done. --Ligulem 23:48, 19 February 2006 (UTC)
If we don't add "Ed." after editor, the field is redundant, as the user may as well just use others. I really think we should break compatibility at some point and add "Ed." or "Edited by", fixing with WP:AWB. ··gracefool |☺ 03:03, 20 February 2006 (UTC)
Uuuh! But I won't do this. So much work for such a small problem. BTW AWB is not almighty because you cannot do everything with regexes. And I'm still waiting for Media-Wiki built-in conditionals. --Ligulem 08:48, 20 February 2006 (UTC)
The reason the 'book reference' template had an editor parameter, I thought, was for when you referenced a chapter, i.e. Bloggs's "Chapter X", in Smith's "A Very Long Book". (This is why editor was next to title, not next to the author parameters.) If you're just referencing the book, then you treated the editor as if he were the usual main author, i.e. Smith, John (ed.), "A Very Long Book". And (separate issue) I used others for translators, same as illustrators, etc. —Serein 17:39, 20 February 2006 (UTC)
We now have a chapter paramater (and even a chapterurl). But despite this, I now think we shouldn't add "Ed." to the editor automatically. We need not being so rigid. Let's leave this as it is. --Ligulem 18:36, 20 February 2006 (UTC)

Update from User:Gracefool/cite book

I've just updated the template and documentation. I probably was a bit bold, but the documentation is much improved, and the template doesn't break compatibility. It adds origdate, origyear, origmonth and language. Language is helpful when the book is not in English (there are cases where people have wanted this and have had to add a note after the template).

Note that one regression test is broken, but it was incorrect usage anyway (they manually linked the date). ··gracefool |☺ 03:40, 20 February 2006 (UTC)

Ha! In use. See this (origyear). And this (language on last book). --Ligulem 09:39, 20 February 2006 (UTC)
I've fixed two regression tests: [1] and [2] (red date). Other ones still need updates too (blue year on manual expansions is missing). Additional test cases for the new features would also be good. --Ligulem 09:51, 20 February 2006 (UTC)
Updated WP:CITET: [3]. --Ligulem 10:01, 20 February 2006 (UTC)
It might be a good idea to check for wikilinking of dates when converting {{book reference}} with AWB. ··gracefool |☺ 12:29, 20 February 2006 (UTC)
Yeah. I'm doing this. It's a real pain though. Needs to be done manually as there are too complex cases like this (edited manually in the edit window of AWB). BTW you could boot up on the AWB camp too :-) (I hope you do have a Windows box somwehere). I think it will take a bit longer to convert to cite book now... --Ligulem 13:10, 20 February 2006 (UTC)
See also User talk:Adrian Buehlmann/AWB/migrate to cite book#Redlinks in year/date. --Ligulem 16:05, 20 February 2006 (UTC)
On Binding of Isaac there was:
* {{Book reference
 | First=Flavius | Last=Josephus | Authorlink=Flavius Josephus
 | Title=Antiquities of the Jews
 | Publisher=Interhack Digital Library
 | Year=93-94 C.E.
 | ID=ISBN 1-58-827612-0
 | URL=http://www.interhack.net/projects/library/antiquities-jews/b1c13.html
 }}
I have converted that to:
* {{cite book
 | first=Flavius | last=Josephus | authorlink=Flavius Josephus
 | title=Antiquities of the Jews
 | publisher=Interhack Digital Library
 | id=ISBN 1-58-827612-0
 | url=http://www.interhack.net/projects/library/antiquities-jews/b1c13.html
 }} 93-94 C.E.
in order not to produce a redlink for the year parameter. The old book reference accepted that. The current cite book does not. --Ligulem 16:16, 20 February 2006 (UTC)
Surely this is less than helpful and should be fixed: we're losing information rather than preserving it :-( —Phil | Talk 16:48, 20 February 2006 (UTC)
NB this is also now incompatible with {{cite journal}} which requires manual linking of dates, just like {{book reference}} used to. I think we should return to the previous situation because this is just too much hassle. HTH HAND —Phil | Talk 17:04, 20 February 2006 (UTC)
Gone back (see also talk below): cite book, Binding of Isaac. --Ligulem 18:29, 20 February 2006 (UTC)

Automatic wikilinking of dates

Sorry for bugging you guys once again; I know all the hard work you do to get this template to work right, and to be efficient at the same time. Regarding the wikilinking of dates, some times I don't think the wikilinking should be automatic. I bring about the Baha'i literature article again (and I know this is a minor problem, and usually is not a problem) in that the original publication dates may span years (as is the case with compilations of various tablets written in the 18th century and before), and thus the date fields look red. -- Jeff3000 17:45, 20 February 2006 (UTC)

And it's often of dubious merit anyway. I'm interested to link to a page on the author, or on the book, or the ISBNs page; personally, I'm less excited to find out what unrelated things happened in the year the book happened to be published! —Serein 18:03, 20 February 2006 (UTC)
I think I'm going to remove that wikilinking of year, month, date, origyear, origmonth and origdate now from the template. This causes more trouble than it adds usability. Please hold the line... --Ligulem 18:14, 20 February 2006 (UTC)
Done. --Ligulem 18:21, 20 February 2006 (UTC)
Thanks, appreciated. —Serein 19:22, 20 February 2006 (UTC)
  • The reason I did it was for automatic date formatting. How about we re-add the autolinks for date and origdate. That shouldn't cause many problems, since those fields are new and require full dates in the right format for wikilinking. ··gracefool |☺ 20:35, 20 February 2006 (UTC)
I agree it makes sense for fields that require full dates, and with those we won't get Jeff3000's red links problem either. OK by me. :) —Serein 21:27, 20 February 2006 (UTC)
No please not. BTW the date parameter is not new. {{Book reference}} already had a Date parameter (look at the template code and the detailed description of the params) and it is used (although sparingly). I find it very laborious to hunt for uses of date and convert the dates to ISO format by hand to avoid stupid redlinks. And this all just to have a link to a date. No, please not. Too much work for too little return on investment. --Ligulem 21:44, 20 February 2006 (UTC)
It needs to be in ISO format? Yikes. I see your point. —Serein 22:02, 20 February 2006 (UTC)
I agree with Adrian on this, wikilinking dates is not worth the effort that is needed to make sure everyone has the dates in the right format. -- Jeff3000 01:42, 21 February 2006 (UTC)

Partial solution

We could autolink dates which are specified as month and year which would fix the preference problem: if we inserted [[month]], [[year]] this would trigger the proper code. If people then wanted to specify non-standard dates, or ranges, they could use date like before. HTH HAND —Phil | Talk 15:29, 21 February 2006 (UTC)

I've lost the whole picture about these dates. Aren't these dates a bit a mess now in the citation templates? For example cite web (old web reference) does now have the same accessdate and accessyear parameters as here, but their semantic seems to be a little different from here (inherited from web reference). For example stuffing of a month and a day of month into "accessdate" without a the year seems a normal use case over there. I'm confused. --Ligulem 15:54, 21 February 2006 (UTC)
On {{cite web}} (derived from web reference) we have:
{{qif
  |test={{{accessdate|}}}  
  |then= URL accessed on [[{{{accessdate}}}]]{{qif
    |test={{{accessyear|}}}
    |then=, [[{{{accessyear}}}]]
  }}
}}
which leads to URL accessed on [[{{{accessdate}}}]], [[{{{accessyear}}}]] if both accessdate and accessyear are specified in the call (and this auto-date-links-preferences-aware if used as accessdate=July 12 and accessyear=1978). This compares to your [[month]], [[year]] proposal for cite book. --Ligulem 16:05, 21 February 2006 (UTC)
On cite book we now have:
{{qif
  |test={{{accessdate|}}}
  |then=. URL accessed [[{{{accessdate}}}]]
  |else={{qif
    |test={{{accessyear|}}}
    |then=. URL accessed {{qif
      |test={{{accessmonth|}}}
      |then=on [[{{{accessmonth}}} {{{accessyear}}}]]
      |else=during [[{{{accessyear}}}]]
    }}
  }}
}}
--Ligulem 16:09, 21 February 2006 (UTC)
What about inheriting names and semantics of accessdate and accessyear from cite web and replace in cite book
{{qif
  |test={{{accessdate|}}}
  |then=. URL accessed [[{{{accessdate}}}]]
  |else={{qif
    |test={{{accessyear|}}}
    |then=. URL accessed {{qif
      |test={{{accessmonth|}}}
      |then=on [[{{{accessmonth}}} {{{accessyear}}}]]
      |else=during [[{{{accessyear}}}]]
    }}
  }}
}}
with
{{qif
  |test={{{accessdate|}}}  
  |then=. URL accessed on [[{{{accessdate}}}]]{{qif
    |test={{{accessyear|}}}
    |then=, [[{{{accessyear}}}]]
  }}
? --Ligulem 16:16, 21 February 2006 (UTC)
What about outright nuking of the month, origmonth and accessmonth parameters on cite book and abusing date and year parameters like in cite web? (Sorry, I hope I can stop my chatter now). --Ligulem 16:33, 21 February 2006 (UTC)
after edit conflict You have a problem now in that you have started me thinking :-)
So whip off that automatic linking for accessdate and stick on an extra layer, which should give us:
{{qif <!-- do we have any date info at all? -->
  |test={{{accessdate|}}}{{{accessyear|}}} <!-- either of these will do -->
  |then=. URL last accessed {{qif <!-- if we have "date", discard everything else -->
    |test={{{accessdate|}}}
    |then={{{accessdate}}} <!-- manual linking to allow for unorthodox formats or ranges -->
    |else={{qif
      |test={{{accessyear|}}}
      |then={{qif <!-- if we have "year" then we need to know if there's "month" also -->
        |test={{{accessmonth|}}}
        |then=[[{{{accessmonth}}}]],&#32;[[{{{accessyear}}}]] <!-- trigger date preferences -->
        |else=[[{{{accessyear}}}]]
      }}
    }}
  }}
}}
We don't need all that silly "on/during" nonsense. We can then shove that code into {{cite web}} (and possibly {{cite book}} if we want to date any URLs there) so everything is working the same.
Do you see what I mean about real meta-templates possibly proving helpful? We could code this in something like {{citation dates}} and ensure that all citation templates used the same format.
HTH HAND —Phil | Talk 16:48, 21 February 2006 (UTC)
(Side note). If you really want to go that road, I would propose to take singular {{citation date}} instead of "citation dates". --Ligulem 08:24, 22 February 2006 (UTC)
No (so now you have a problem :-). I propose to do the following:
  • Nuke month, origmonth and accessmonth. We simply need no months. Have you seen Month set on any call of Book reference? Nobody uses that.
  • No date linking on date, year (we simply don't need that)
  • No date linking on origdate, origyear (we simply don't need that)
  • Same handling for accessdate and accessyear as in {{cite web}}
So I would do the following code for cite book: User:Adrian Buehlmann/work/cite book/2006-02-21
--Ligulem 19:23, 21 February 2006 (UTC)--Ligulem 17:47, 22 February 2006 (UTC)
Sounds good — but that means no user-preference dates for original date or publication date :( ··gracefool |☺ 23:31, 21 February 2006 (UTC)
Correct. But we keep the flexibility as it is currently demanded from existing calls on the year and date paramaters and we do not pile up another level of (maybe?) brittle code (qif is still in danger). I have also some bad feeling in my stomach about encoding the date-wikilinking into the (month,year) call pattern. The wikilinking fans of the editors will start inserting dates using this method. This causes an obscure segmentation of main usage patterns. But if Phil insists on that I do not want to play the progress killer here. I know that it hurts when people reject proposals. Oh well. Maybe I have a bit a wrong view on this. I'm not that sure on this issue. And I apologize for being so verbose here. --Ligulem 08:17, 22 February 2006 (UTC)--Ligulem 17:47, 22 February 2006 (UTC)
Ok, I don't care anymore because it looks like all dates will be auto-formatted in future - this is bug 4582. Go vote for it. ··gracefool |☺ 09:19, 22 February 2006 (UTC)
I withdraw my proposal to remove the month params. It's too much work to go through all calls and check that months are not used. I can't do that. Not enough wiki time. --Ligulem 17:47, 22 February 2006 (UTC)

Usage examples

Some actual usage examples:

  1. year=2003: [4]. Very common.
  2. date=[[May 18]], [[2004]]: [5], [6] . Manual wikilinking, prefs enabled.
  3. year=1849, 1973: [7], [8]. Two years. Could be manually converted to origyear.
  4. year=1929-1991: [9]. Year range?
  5. year=1997 | month=December: [10], [11], [12]. month is used!
  6. year=2001 | month=August 9: [13], [14]. month and year used, with day of month and month stuffed into month parameter.
  7. year=1985 originally 1966: [15]. Another two year variant. Could be manually converted to origyear.
  8. year=November, 1991: [16], [17], [18]. Stuffing month and year into year parameter.
  9. year=[[1982]]: [19]. Manually linked year.
  10. year=2 edition (December 18, 2003): [20]. Stuffing edition and date with formatting into year. Rare abuse case.

--Ligulem 16:13, 25 February 2006 (UTC)

Punctuation

There needs to be some punctuation, probably a full stop after the name and/or year, so that it renders in the same way as {{book reference}}. At present, it is absent. I can't for the life of me work out where to put how many full stops, but it needs doing fairly urgently given the migration that is already underway. -Splashtalk 21:39, 21 February 2006 (UTC)

Could you give at least one example where cite book renders differently compared to book reference? I have seen many renderings of cite book, also on the regression test page. But I didn't miss any full stop. What have I overlooked? --Ligulem 21:47, 21 February 2006 (UTC)
It turned up in article I'm writing, so I can't show you in place, but here is the particular usage verbatim:
With {{book reference}}:
John G. Proakis (1995). "14" Digital Communications, 3rd edition, 767–768, Singapore: McGraw-Hill Book Co. ISBN 0-07-113814-5.
With {{cite book}}:
John G. Proakis (1995). “14”, Digital Communications, 3rd edition, Singapore: McGraw-Hill Book Co, 767–768. ISBN 0-07-113814-5.
Note the missing full stop after (1995) in the latter. Incidentally, this raises the fact that the rendering of chapter is pretty weird. -Splashtalk 03:32, 22 February 2006 (UTC)
Fixed missing period. I've clarified the role of chapter in the documentation - your examples aren't using it correctly, they should be "Chapter 14" instead of "14". It's really designed more for chapters that have titles, rather than just meaningless numbers. ··gracefool |☺ 07:29, 22 February 2006 (UTC)
Thank you. And thanks for clarifying the intended use of chapter, too. -Splashtalk 16:04, 22 February 2006 (UTC)

Multiple authors

Isn't it normal to have no comma but an "and" before the name of the final author? I notice this is missing at present. -- Jheald 15:29, 22 February 2006 (UTC).

The template doesn't add "and", you have to write it yourself. Makes sense really, because it depends on how many authors you add. —Serein 16:01, 22 February 2006 (UTC)

Demarcate co-authors with semi-colon

I'm busy migrating {{book reference}} to {{cite book}} and I'm finding it difficult sometimes to distinguish author names when multiple authors have so many commas. I've been chopping out second and subsequent authors into coauthors and converting the first author to first and last, but this often results in silly-looking lists with commas and semi-colons scattered over the shop. This was not helped when the primary author of one particular book seems to be referred to by just his last name, which rendered the list out-of-step. I would like to propose that we move towards separating the (co-)authors with semi-colons, making the commas which separate first and last more easily distinguishable. As a first move, I would like to alter this, and related templates, so that coauthors is preceded by a semi-colon rather than a comma. Thoughts?

—Phil | Talk 14:05, 23 February 2006 (UTC)

Or abandon co-authors. I never use them when I use either this template or cite journal/journal reference. Does the use of them follow some scheme or somethgin? -Splashtalk 01:14, 24 February 2006 (UTC)
No. We need coauthors. Param author is deprecated. If we then have only first and last, where should we add additional authors? The first author is special on books as his last name and the year together form an index to the book. --Ligulem 06:50, 24 February 2006 (UTC)
Sorry, what? Author is deprecated? That's just madness. -Splashtalk 13:23, 24 February 2006 (UTC)
You might want to talk with gracefool. He deprecated that. I don't care. --Ligulem 13:40, 24 February 2006 (UTC)
For primary author, author was deprecated in favour of first and last. For coauthors, coauthors should be used instead (not deprecated author or author2, author3 etc.). Yes, coauthors do follow some style, but which one you use, if you use one at all, is complicated. —Serein 13:46, 24 February 2006 (UTC)
Support. --Ligulem 06:47, 24 February 2006 (UTC) per Serein --Ligulem 13:38, 24 February 2006 (UTC)
A semicolon before coauthors doesn't adhere to Chicago or Harvard citation styles. (The comma currently in place already breaks MLA or APA, as far as I'm aware.) Given we're not recommending a particular style, should we be mandating a new, otherwise unsupported citation style? —Serein 13:35, 24 February 2006 (UTC)
I went to look at the originals of the various citation styles listed there, and discovered that the guy who hosts three of the five descriptions has created his own style which…guess what?…uses semi-colons to delimit co-authors . So far as I can tell, the overriding necessity is for the information to be presented clearly and unambiguously, and all those stupid commas get in each other's way. HTH HAND —Phil | Talk 16:58, 24 February 2006 (UTC)

Use with EndNote

It is perfectly possible to make an Output style for the EndNote reference management program (and no doubt other, similar products) to handle all (or almost all) of this automatically on a editor's own computer, which can then be pasted right into Wiki as needed. There is one major hang up however, and that is the author's name. EndNote (and similar) does not have separate fields for first and last name, nor does it put multiple authors in separate fields. In this regard the old param author was superior. Any suggestions on how to deal with this or is it just a matter of editing the output? Klompje7 10:06, 24 February 2006 (UTC)

Just drop the information into the author field and one of our janitors will be along to clean it up later. HTH HAND —Phil | Talk 16:59, 24 February 2006 (UTC)

How do I enter authors with more than two names?

I'm completely baffled as to how to use this template for at least half the references I need to enter.

Obviously for "John Steinbeck" I use first = John, last = Steinbeck," but just how am I supposed to enter

  • H. G. Wells
  • Albert Payson Terhune
  • J. R. R. Tolkien

Just use the authorlink and leave out the first and last? If so, then what if the author has no Wikipedia article?

What do I do when there is more than one editor?

What format, exactly, am I supposed to use for coauthors? Dpbsmith (talk) 10:49, 24 February 2006 (UTC)

  • first=H. G. | last=Wells
  • first=Albert Payson | last=Terhune
  • first=J. R. R. | last=Tolkien
Add the second and all other authors to coauthors in the form "First Last, First Last, First Last". You can also take a look at the test page at Template:Cite book/regression tests. --Ligulem 11:02, 24 February 2006 (UTC)