[mlz] new style coming up

My twlaw.csl is now in a usable condition, with me and my students alpha-testing it. What should I do next to have it listed on the MLZ site? Or should it go to Zotero's style repo?

I also wonder if you may consider expanding CSL's dependent style/locale file capacity, to allow linked styles. Take my twlaw.csl as an example, which currently contains 2 locales: English (default) and Chinese. The English part is taken verbatim from mlz-amlaw.csl, as the Bluebook is now the de facto standard for English citation here, despite the lack of standard for Chinese citation. When amlaw is updated, so is twlaw. It would therefore be much more convenient and less error-prone if I may simply set my twlaw style up as an adjunct style to amlaw.

In the long run, I think the best approach for a truly multi-lingual environment would be to allow the user to decide--for citation and bibliography respectively--which style to use for each language/locale. Personally I need only citation styles for English & Chinese material. If MLZ does catch on, however, I'll probably get requests for German, Japanese and French styles sooner or later. Trying to maintain consistency will be a challenge.

This per-language/locale style preferences can be done either through the UI or, if that's too complicated, by allowing something like <include mlz-twlaw.csl/> to be used inside <layout></layout>.

Thoughts for the future. For now, I'm just happy to have MLZ. It's a godsend.
  • Great stuff!

    Modularity is indeed the most important remaining issue for the CSL-m extensions, both for language domains and jurisdictions.

    Allowing explicit dependencies in the CSL itself would be complicated, since we would need to assure that the called styles are accessible, and compatible with the calling style. It would also complicate debugging work, and would require that the maintainer of the "top of the food chain" styles splice in additional jurisdictions and language locales on demand.

    I've been thinking that it may be possible to build styles on demand through a "smart" distribution site. A user would choose their "core" style, and then add jurisdiction and language modules to it by making selections from a GUI. The distribution server would compose the style XML and deliver it as a single file.

    That would take some work to set up, and it would require significant effort to work out a consistent "API" for modular style elements that plays well across known style structures -- back-referencing for Commonwealth case cites, say, should work smoothly when imported for use in US Bluebook, for example, without major refactoring.

    It would require serious server-side skills, and participation by multiple parties familiar with the demands of cross-jurisdiction and cross-language style design. It should be a funded effort, at least for the first burst of development work. I haven't pushed hard in that direction yet: I did apply, with a focus on law, to one of the foundations that has supported Zotero, and was told that they don't fund CSL projects "for specific fields". But that was just one attempt.

    Let's keep in touch on this: it would be great to start working through the details.

    The MLZ styles are in a GitHub repo. If you send me a private Inbox message via the forum, we can get you squared away on the update workflow, and get the style up on CitationStylist.
  • For the interim, and along the same lines, there is a maintenance script in Python that I used for keeping the initial six styles "in sync". With it, you can specify a "parent" style and one or more macros to import. The script can remove the macro and all macros that it calls from the style, and replace it with the macro chain from the parent.
  • Private message sent. And thanks for the script. I'll certainly need it down the road.
  • I've made a very small cosmetic change to mlz-twlaw.csl on the CitationStylist repository. The change replaces a few numeric entities with their UTF-8 character equivalents, for consistency.
  • Got it. Thanks. But that makes me wonder: do I need to do the same to the various Chinese terms/text in twlaw?
  • And oops ... Emacs nxml still declares the style invalid. I had missed a setting while testing, and it wasn't actually applying the schema.

    The online validator works though, and isn't phased by the Chinese characters.
Sign In or Register to comment.