Hi Pete,
You are correct.  The new schemas need to be reloaded.  It's on my list
of things to do, but will likely be sometime in January before I can
look at it.

Cheers!
h

-----Original Message-----
From: DCMI Registry Community [mailto:[log in to unmask]] On
Behalf Of Pete Johnston
Sent: Thursday, December 21, 2006 5:09 AM
To: [log in to unmask]
Subject: DCMES data in DCMI Registry

I'm not sure whether to send this here or to the DCMI Web site admin
folk, but as it raises some broader issues, I think it's probably worth
sending it here.

This week the content of the "namespace document" describing the terms
of the DCMES was updated to reflect changes made by the Usage Board. See
news item

http://dublincore.org/news/documents.shtml#dcmi20061218terms

However, I notice that the views of the descriptions of the DCMES terms
provided by the DCMI Registry are still based on the old data e.g.

http://dublincore.org/dcregistry/detailServlet?reqType=detail&item=http%
3A%2F%2Fpurl.org%2Fdc%2Felements%2F1.1%2Fsubject

> Definition: The topic of the content of the resource. 

cf. http://purl.org/dc/elements/1.1/subject

i.e. the DCMI registry database has the old RDF data stored. I guess
someone needs to "crank the handle" which gets it to reload the data
from the "namespace document", which is the authoritative source. 

I guess the same is probably true of various other "metadata registries"
out there.

And I guess some thought needs to be given as to how that mechanism
works in a distributed environment i.e. the DCMI registry is only
reading data provided by DCMI, but other such applications may be
reading data sources from around the Web, owned by many different
agencies. 

Is it the data source owner's responsibility to ping a registry to say
the data source has changed and the registry should reload/reindex? What
if the data source owner doesn't even know that some registry has loaded
their data? (Could data source owners provide RSS/Atom feeds to do
this?) 

Or does a registry periodically re-GET all its current data sources to
check for changes from the version that was stored? Are there other
implications of doing that? 

I seem to recall Harry did some work on mechanisms whereby multiple
instances of the DCMI Registry could exchange data, though I'm afraid I
can't recall the details. But in any case, that only addresses
dissemination between those systems, not how data gets into those
systems in the first place.

Pete

---
Pete Johnston
Technical Researcher, Eduserv Foundation
Web: http://www.eduserv.org.uk/foundation/people/petejohnston/
Weblog: http://efoundations.typepad.com/efoundations/
Email: [log in to unmask] 
Tel: +44 (0)1225 474323