bibxml reference files no longer show all the name information

[cabo]

Describe the issue

https://github.com/ietf-tools/bibxml-service/issues/280

Only fullname seems to be filled in any more. surname or initials seem to be missing now. This creates incorrect guesses as to what the surname and what the initial-generating names are.

This seems to have happened around August 18 to 20

Code of Conduct

• [X] I agree to follow the IETF's Code of Conduct

[rjsparks]

Bringing the rest of the discovery into this issue (pulling text from https://mailarchive.ietf.org/arch/msg/tools-discuss/4C8uDbOjs-o97ivlKfDlA2F33HI>):

This didn't change recently - it's been this way for years, but it is a bug.

https://datatracker.ietf.org/doc/bibxml3/draft-ietf-tsvwg-l4s-arch.xml

and

https://datatracker.ietf.org/doc/bibxml3/draft-ietf-tsvwg-l4s-arch-20.xml

return different results (specifically around author name decomposition), which is not expected. (Nothing special about that draft - this same behavior will occur with any draft).

This bug is rooted in the code-path passing the Document object to the template in one case, and the most recent DocHistory object in the other, and these things have different knowledge of the Submission object that led to the version. It's not clear to me why we have this distinction in this particular code flow and that the template cares, and I'll look to see if it can be simplified.

I'm speculating that if we have the submission object there's more trust in the freshness of the author names, but why we grab initials and lastname in one path and not the other is very much not obvious.

That said, there is rising awareness that trying to calculate initials and "last name" doesn't work across all our contributing cultures, and in the long run, we may stop trying.

[bbriscoe]

Thanks @rjsparks.

Despite the apparent desire to move from "surname, initials" to "fullname" in references, that's a policy question, which may take some time working through "the committee stages".

So, yes, it's important to fix this mechanism problem so that the current "surname, initials" policy works. Because I'm sure some people with names that don't conform to the assumptions will have tried to lever their name into these fields in a way that produces their desired outcome under the current policy. Thank you for confirming that you're making progress on that yourself (with the help of the clues that Carsten unearthed). Someone knowledgeable taking responsibility was the part that was previously missing.