> Jerry McDonough has completed a new version of the METS schema.
> The new
> version is is available at:
> and is ready for review. Please submit comments to the METS listserv.
The new schema has comments on specific changes in the
comments at the top. The major change is the addition
of a new section for METS object behaviors, courtesy
of Sandy Payette at Cornell and the Fedora gang.
This seems a good time to report on some issues that
came up at the Pittsburgh meeting. One question which
arose was whether the use of ##any element declarations
on the mdWrap element could be modified to allow
for more than one root element beneath mdWrap. It
turns out that this is possible. However, doing
so means that the various root elements can come
from different namespaces/extension schema. My
personal feeling is that this is, if not bad, at
least not elegant. I would rather have the elements
within any given mdWrap all come from a single extension
schema. However, I can see benefits for greater flexibility
here as well. I think we should try to discuss this
on the list; if the consensus is that opening up
mdWrap for multiple schema subelements is a good idea,
I'll make that change.
Another issue that arose was whether it was possible
to alter the METS schema to allow for attributes from
other namespaces within METS elements. While this is
possible, it turns out that the changes necessary to
enable foreign namespace attributes on METS elements
compromise validation software's ability to ensure
ID/IDREF matching on any element where attributes from
other namespaces are allowed. While there are potential
work arounds, they are rather ugly. For the moment,
I have decided to forego enabling foreign namespace
attributes in METS.
Hope everyone has a happy holidays!