thinking about conformance

As I’m working on the python implementation, and as we’re trying to
wrap up the remaining tickets, it occurs to me it might be worth
thinking about levels of conformance. Not exactly sure how to do that,
or how formal to be, but am just noting there’s a big difference
between:

  1. formatting a bibliography

  2. 1 + basic citation processing

  3. 2 + complex citation disambiguation

  4. 3 + one of the features Frank and I are talking about (note back
    references and such)b

Bruce

Not exactly sure how to do that,
or how formal to be, but am just noting there’s a big difference
between:

Seems like a good idea, although how does your list relate to the
ability to create papers/reports with correctly formatted citations
for various journals or other outlets? In other words, if a CSL
implementation does just (1) or just (1+2), would it actually be
useful to end users or only as a milestone for developers?

Regards,
Robert.2009/7/31 Bruce D’Arcus <@Bruce_D_Arcus1>:

I guess the simplest way to do this would be to categorize the test suite
(in ‘core functionality’, ‘basic citation processing’, etc), plus some
documentation on how the categories are defined.

Rintze

Not exactly sure how to do that,
or how formal to be, but am just noting there’s a big difference
between:

Seems like a good idea, although how does your list relate to the
ability to create papers/reports with correctly formatted citations
for various journals or other outlets? In other words, if a CSL
implementation does just (1) or just (1+2), would it actually be
useful to end users or only as a milestone for developers?

1 can only generate reading lists and such, so more the latter.

2 can only support simple numeric or label styles in the sciences.

You need 3 for author-date styles.

4 will be needed for note styles and bluebook.