Upgrade StorageA mechanism for handling variations within styles?
Short version: Would it make sense to find a way so that common variations in style output could be handled without forking a CSL style, but by user-specifiable options within the style itself?
Longer version: I can think of a number of small things within the Chicago style which could be said to be discretionary. Presumably the same would be true of any of the 'mother' styles which are used across disciplines for a variety of purposes. I have a written list that I've been collecting, but here are a few off the top of my head:
1. Default to Journal Abbreviations vs. default to full Journal Names
2. Default to Series Abbreviations (when they get added to the UI/Database) vs. full Series Names
3. Never include URL for book or Journal Article, (I know this should really be solved another way, but a the moment, my database is full of imported URLs which I don't want in citations).
4. Use long initial citations AND print a bibliography (as required for many theses/dissertations), and discussed in the last three comments here, and gerv's comments here.
5. There may be other issues with date format or capitalization of non-English titles, etc. I can't think of other examples, just now. People ask on the "Which citation formats?" thread of this forum for things like Turabian (or Society of Biblical Literature), both of which are based on CMS, but may have some slight variations which I haven't discovered yet.
It seems like there are two options for dealing with discretionary (or other small) variations in the canonical styles. (A) Make a flexible CSL style, which would include 'options' that the user somehow had access to, or (B) duplicate and modify styles as needed. (B) needs no other infrastructure, but might be inefficient in the long run because everyone maintaining a variant for themselves (or a community) needs to merge changes from the main style every time it is one is improved (or CSL gets upgraded). If variants turn out to be common, it could also lead to a proliferation of slightly differing styles. (A) sounds very nice to me, especially for common variants, but it would need support for options in CSL and UI support in Zotero et al. It would of course still be possible (and sometimes required) to fork a style for individual requirements. Any thoughts on this?
Longer version: I can think of a number of small things within the Chicago style which could be said to be discretionary. Presumably the same would be true of any of the 'mother' styles which are used across disciplines for a variety of purposes. I have a written list that I've been collecting, but here are a few off the top of my head:
1. Default to Journal Abbreviations vs. default to full Journal Names
2. Default to Series Abbreviations (when they get added to the UI/Database) vs. full Series Names
3. Never include URL for book or Journal Article, (I know this should really be solved another way, but a the moment, my database is full of imported URLs which I don't want in citations).
4. Use long initial citations AND print a bibliography (as required for many theses/dissertations), and discussed in the last three comments here, and gerv's comments here.
5. There may be other issues with date format or capitalization of non-English titles, etc. I can't think of other examples, just now. People ask on the "Which citation formats?" thread of this forum for things like Turabian (or Society of Biblical Literature), both of which are based on CMS, but may have some slight variations which I haven't discovered yet.
It seems like there are two options for dealing with discretionary (or other small) variations in the canonical styles. (A) Make a flexible CSL style, which would include 'options' that the user somehow had access to, or (B) duplicate and modify styles as needed. (B) needs no other infrastructure, but might be inefficient in the long run because everyone maintaining a variant for themselves (or a community) needs to merge changes from the main style every time it is one is improved (or CSL gets upgraded). If variants turn out to be common, it could also lead to a proliferation of slightly differing styles. (A) sounds very nice to me, especially for common variants, but it would need support for options in CSL and UI support in Zotero et al. It would of course still be possible (and sometimes required) to fork a style for individual requirements. Any thoughts on this?
I will say that whatever solution needs to strongly, strongly encourage style be available for sharing. We have to stop thinking about individual user styles. The goal here is to create an infrastructure such that users don't have to think about styles, other than to choose them.