Should there be a: No given-name disambiguation default style?

2»
  • dda-gre
    Frank's suggestion worked so well for me that I just tweaked the AGU style into a copernicus one, that I will send to the repository... as soon as I get the last little glitch solved...

    I need the bibliography to read like that:

    Mouse, M., Duck, D., and Dysney, W.: A rollercoaster ride to remember, Disney Magazine, 666, 19-31, 2005.

    except it reads :

    Mouse, M., Duck, D., et Dysney, W.: A rollercoaster ride to remember, Disney Magazine, 666, 19-31, 2005.

    it has to be in this author macro (actually almost the one from the agu.csl):

    <names variable="author">
    <name sort-separator=", " initialize-with=". " name-as-sort-order="all"
    delimiter=", " delimiter-precedes-last="never" and="text"/>
    <label form="short" prefix=", " suffix="." text-case="capitalize-first"/>
    </names>

    is it because I'm french that the "and" text reads "et" ?

    I thought the xml:lang="en" on the first line of the .csl was setting the locales !
  • adamsmith
    edited August 11, 2009
    no, the locales are set according to the hidden preferences in Firefox
    http://www.zotero.org/support/hidden_prefs#export_and_citation_settings


    P.S. before you get frustrated with SVN let us know and we'll either help you and/or upload the style for you.

    @Frank - ah, come on, disambiguation is you secret passion, admit it, you're among fellow nerds here.
  • dda-gre
    fantastic, it works just fine now...

    the new copernicus style is ready to upload !
  • dda-gre
    fantastic, it works just fine now...

    the new copernicus style is ready to upload !
  • brazuca
    @adamsmith and @fbennett:

    Indeed, our style guide requires the initial only for disambiguation purposes. So we are back to not having this capability in the current version. It is not a huge problem in the short run, but it would be hard to justify pushing Zotero in my organization without it.

    Are there plans to incorporate this ability? It would seem that there is a need for greater control over the style of the disambiguation layout and rules.
  • adamsmith
    yeah - I looked into the
    <if disambiguate="true"> option, but that seems weird.
    Everytime you select multiple items it kicks in?
    i.e. by putting

    <choose>
    <if disambiguate="true">
    <text value="test1"/>
    </if>
    <else>
    <text value="test2"/>
    </else>
    </choose>

    either in citation or bibliography (of any given style) I get "test1" whenever I select two citations and test2 whenever I select only one.

    Is that intended?
  • dda-gre
    Hi,

    I have been tryin to make sense of SVN... and It's just not going to happen...
    @adamsmith: you suggested that you could direcly upload the style for me...

    that could be an idea: other wise, I'll leave on vacation friday, and totally forget about it at my return...

    I copy the style directly here, or is there some less messy way of doing that ?

    dda
  • fbennett
    edited August 12, 2009
    @brazuca,

    Yes, this will certainly be supported! Work on an upgrade of the CSL specification from version 0.8 to 1.0 is near completion, with the target for completion just two weeks away. It will be accompanied by a new CSL processor that will support the higher specification. The processor will be part of a large revision to the plugins and related machinery in Zotero, which will take some more time to complete.

    The new version of CSL will support seven separate disambiguation algorithms with subtle differences between them, and should cover all known requirements.

    There's no deadline for fnal deployment of the new processor, but the work is definitely in progress.
  • fbennett
    @dda-gre,

    If you're online now, we might give this service a try: http://www.jetbytes.com/

    You need to keep the page open in your browser with your file queued for upload, and post the link here. I'll click through and pick up the file.
  • brazuca
    @fbennett

    Great news then! I'll just wait for it when it comes out.

    PS: will the new CSL break existing styles?
  • fbennett
    @brazuca: There will be some backward-incompatible changes, but the language designers plan an automated upgrade of styles in the repository. You should be able to just carry on working.
  • adamsmith
    dda-gre:
    I always post to
    gist.github.com and post here - no registration, short links etc.
  • fbennett
    @adamsmith, dda-gre: Ah, that's a better plan.
  • dda-gre
    great ! that was an easy one !!!

    I'am not sure which address is really valid, so I put both...

    here is a proposition for a csl style for copernicus published journals (Atmospheric Chemistry and Physics, Climate of the Past, The Cryosphere, Ocean Science, ...)

    it is based on the csl for agu journals, found in the zotero repository.

    git://gist.github.com/166857.git

    http://gist.github.com/166857

    vacations are approaching !
    thank yall for your help
  • adamsmith
    it's up. thanks for contributing!
  • komrade
    Frank,
    Just to confirm re our discussion of June 13, Chicago MOS replied to my query and say that yes, Chicago disambiguate primary author only, and the initial separator is ". "

    And when you get a chance could you post that standard list of references which would demonstrate the correct formatting and disamb'? I'll be building some new styles when CSL 1.0 is ready and I'd like to use a standard list, then when I think it's done, send it to the publishers to confirm that it's all correct.
    Cheers!
  • fbennett
    Thanks for the reminder; I'll add that to the todo list. If it's to be used with CSL 1.0, it can be cast as a set of standard test fixtures. That way the test runner will tell you where the errors are, and you can save some time trawling through the output.
  • komrade
    edited August 31, 2009
    A little light bedtime reading of the CSE manual made me realise that I misread their info on disambiguation.
    CSE does generally specify minimal disambiguation - Frank's behaviour a).
    But if both authors of a two author paper have the same surname, then initials should be used...
    (Smith TL and Smith UV 1990)

    So all 3 behaviours are specified in various style guides - a) minimal, b) primary author, and c) all authors, and in CSE, a slightly weird hybrid.

    Sorry for the previous misinformation - I'm glad you kept it in the CSL!
  • fbennett
    edited August 30, 2009
    Yay! Uh-oh. This is another permutation, unfortunately. If there are no confllcts with other cites, this would come out as Smith & Smith. But I'll open a ticket for it, and try to work out a way to squeeze it in.
  • komrade
    Well it's a ridiculous rule really. I'm not sure of the rationale. But that's what it says. I read it properly this time.
    (I'll be doing 3 CSE styles when CSL 1.0 is ready)
  • fbennett
    @komrade,

    I've added this to the "by-cite" disambiguation rule (the minimal one). I've added a test, and it seems to work. Thanks for reporting this, hope it works when we turn on the switch!
  • komrade
    Hi Frank
    Looks good for CSE but I can't see in the test code how you would turn it off?

    Can the (Smith JB & Smith RM 2002) behaviour turn on separately from the "by-cite" behaviour? Although it's specified in the CSE guide, I still often see (Smith & Smith) in many Life Science texts.
  • fbennett
    Could you check to see if those Life Science texts use disambiguation at all? It's possible to control this one-cite disambiguation separately from the "minimal" form, but it would require an addition to the schema. We've just put in a set of options meant to cover all known possibilities, so if this is needed, it should be possible to add it at this point, before CSL 1.0 is finalized. But there'll need to be evidence that it's required to round out the coverage. Sorry to put you to the extra trouble, but if you could check that "Smith & Smith" cites and names-based disambiguation (i.e adding initials or names, rather than flaking out with 2000a) do coexist, it would be very helpful.
  • komrade
    Yeah, that's a good point.
    While looking for an example (I searched for Ehrlich & Ehrlich) I quickly realised that it would be very rare indeed for someone to have both a (Smith & Smith) AND a (Bennett FG 2004) and (Bennett JB 2004)
    (i.e. single author, same surname, same year) in the same document.

    I guess just turning off given-name-disambiguation completely in this instance is good enough for all but the very rarest of possible cases.
  • komrade
    But while I was looking, I did find an example a bit the other way:
    i.e.
    (Lee & Lee, 1990) (S. Kim, 1995) (B. Kim, 1990) (S. W. Lee, 1990)

    Thus given-name disambiguation by first author, but (Kim & Kim) without initials.
    I assume the current "first author only" will handle this just fine.
  • fbennett
    edited September 5, 2009
    Yikes. Not quite clear what rule is working there. The first Lee in the Lee & Lee cite should be expanded, in that case. Odd. But as you say, we're down to fine distinctions and rare cases at this point. Some of the examples from the wild that won't fit in the boxes may be due to editorial slips, or just be idiosyncratic choices ("everyone knows that book as Lee & Lee") that would have to be done by hand anyway. Shall we stand pat with what we have in place, and play it by ear?
  • komrade
    Well, if it were possible to separate out the 'co-authors with same name' initials behaviour from the more general disambiguation behaivour without too much pain, then it would I guess allow both of the above cases.
    But if it's unwieldy, just leave it out - I'm sure it would come up so rarely that it really isn't an issue. As you say, editors probably don't have any rules for this, cos it's so rare they just deal with it on gut feeling (and manually) if it ever occurs.

This is an old discussion that has not been active in a long time. Instead of commenting here, you should start a new discussion. If you think the content of this discussion is still relevant, you can link to it from your new discussion.