Ross Singer wrote:
> Agreed.  The same is true, of course, of MARC and, by extension,
> MARCXML.  Part of the "format" is that it can be one record or
> multiple.  I don't think this a particularly strong argument against
> using the namespace as an identifier.
>   
>

Actually, the MARC format (not MARCXML) is very much a single-record 
format. There is a standard for "tape headers" but no wrapper for MARC 
(Z39.2) records, since the MARC format doesn't have a way to do that. 
Having worked for way too long with MARC, I had a lot of trouble with 
the "collection" concept in MARCXML and MODS, and am still not sure I 
see the utility of it beyond what a file of records provides. I'm 
assuming its main purpose is to provide valid XML when you have a file 
with more than one bibliographic record. However, it seems that the 
collection and the records within the collection are part and parcel of 
the same schema, making the things we think of as "records" subordinate 
to the collection, even if it is a collection of one.

kc

-- 
-----------------------------------
Karen Coyle / Digital Library Consultant
[log in to unmask] http://www.kcoyle.net
ph.: 510-540-7596   skype: kcoylenet
fx.: 510-848-3913
mo.: 510-435-8234
------------------------------------