Closed jcolomb closed 2 years ago
move the PID discussion to #20.
This issue was meant for internal ID for authors.
PR discussion:
Do we really need these affiliation IDs? I mean, institutions can have IDs just as ORCID for authors, right? If so, we could fall back to such arbitrary IDs only in case where an institution does not have an ID, and then we could create such an ID ourselves, or?
That refid is only used to link authors with affiliation and created internally. It is needed if we keep affiliation and other author information separate and that the link (author-affiliation) cannot be derived from where the affiliation information is entered. In practice, authors often share affiliation and writing the affiliation in the author information means any tool need to check for duplicates to link each affiliation only once. Also affiliations ID only link to the universities, not to the department, so 2 different affiliations can have the same institution ID.
FWIW, see the following issue for an overview of the current metadata "schema" supported by pandoc (JATS only): jgm/pandoc#6685
at the moment, author internal ids does not seem to be useful.
Prototype at #17 has no id (or anchor) for authors, but has one for affiliation. in #13, it seems we would need some (?), @tarleb mentioned also that a numerical author index would do.
Could we merge the author id element with the author byline order then (see #18) or would this be a very bad idea (as I suspect)?