Quoting Bernhard Eversberg <[email protected]>:
Important question: How best to record these values (code or represent these properties) in the RDA context? And are there new properties/values not yet present or derivable in MARC? (Important things not included in the test records?)
I haven't yet done a comparison of MARC and RDA lists, but I think I have what I need to do so.
(Considering that we need repeatable fields, RDA records can of course not consist of a flat string of data elements on one level. XML permits nesting of elements, down to arbitrary complexity, but do we want that? This complexity would propagate into all software that will have to deal with the data, and don't overlook the input and editing interface and its demands on the cataloger.)
I'm not sure that "flat" and "repeatable" are in conflict. Keeping a set order is the trick, I think, and we should re-evaluate how important order is between repeatable elements. Order of elements in general can be a function of the display application rather than the carrier if the data descriptions are granular. (BTW, RDF does have mechanisms for ordering sets of elements, but it is discouraged for exactly the reason you give above: it adds complexity for users of the data.)
My gut feeling is that in terms of structure of the data carrier we want to impose as little as possible in order to allow for flexibility (perhaps mainly in non-library uses of the data). We should concentrate on the meaning of the elements, and not expect the structure of the carrier to provide that.
kc -- Karen Coyle [email protected] http://kcoyle.net ph: 1-510-540-7596 m: 1-510-435-8234 skype: kcoylenet
