Thanks, Mark, for the translation.
On 7/11/14, 11:16 AM, Mark K. Ehlert wrote:
> The abandoned "info" URI effort leaves me skeptical that non-HTTP URIs
> can be systematically described in general. I'm also skeptical that
> individual identifiers of any kind need to be described inline with
> instance data.
1) what is the use case for describing these identifiers with more
detail than we give them in MARC today?
2) if they do require additional description, there is clearly some
information that is the same for all identifiers of that class (e.g. GPO
numbers -> created by GPO; have a certain format; etc.). This would
argue for treating those identifier types as classes, and putting
instances of the class in the bibliographic data.
OK, three thoughts:
3) as for needing date of creation, that's not particular to this data type.
[log in to unmask] http://kcoyle.net