catching back upů
On Oct 1, 2013, at 10:26 AM, Leif Andresen <[log in to unmask]> wrote:
> The question about Holdings as a class of Annotation or something side-by-side to Annotation is for me not important. The important issue is that Holdings is much more complex than described in the current version. May be it will from a technical view make sense to make Holdings separate?
Holdings are indeed far more complex than the BIBFRAME Annotation Model Draft 2 reflects. But to be fair, this draft isn't designed to define Holdings. The reason Holdings are being discussed in the Annotation document is they echo a "pattern" with other kinds of descriptive requirements that reflect a shareable, abstract representation of a local system's processes or characterization associated with a resource. (Note: a "resource" here may (as you suggested) be a Collection.)
From a BIBFRAME perspective, a Holding is a Class of resource (which means its its own thing). It is also currently underspecified  so, in short, there is lots of room for improvement ;). The BIBFRAME use case document reflects a couple of related use cases , but is not comprehensive and has not been aligned yet with the vocabulary. And while there are many moving parts to consider, the good news here is there is precedent to build on and the requirements you've provided help considerably. But, as is reflected in your and Juha's collective emails, far more work on this particular Class is certainly required.
So in short, keep the feedback coming and thanks in advance! ;)
President, Zepheira "The Art of Data"