Quoting Dan Brickley <[log in to unmask]>:
>
> I took Karen to be referring to mechanisms like Whois, rather than
> indicators of the responsible party that appear in the domain name
> directly.
I was thinking both, actually. There will be some metadata that is
logically tied to a single organization (LC subject headings, for
example). But in general I agree that it is a good idea that a
metadata scheme have its own identifier so that it can move, as Tom
mentioned, as different agencies perhaps take up maintenance duties.
As for caching, it needs to go hand in hand with versioning and with
some agreement about how changes will be handled. Perhaps one of the
possible queries to a URI would return some standard response about 1)
maintenance agency 2) maintenance rules (starting to sound very
ARK-ish, no?).
kc
BTW, the Dewey Summaries project has an interesting challenge in the
area of versioning, and might be a good test case...
http://www.worldcat.org/devnet/wiki/DeweyInfoTechOverview
--
Karen Coyle
[log in to unmask] http://kcoyle.net
ph: 1-510-540-7596
m: 1-510-435-8234
skype: kcoylenet
|