Re: Issue: choice of namespace URI

[James Clark <jjc@jclark.com>]

> / "James Clark" <jjc@thaiopensource.com> was heard to say:
> | > Note also that OASIS can now officially hand out URNs of the form
> | > urn:oasis:... and/or urn:xmlorg:...
> |
> | What allows them to do that?
>
> Informational RFCs; actually now that I go searching for them, they're
> still IDs, but I did receive word from the IESG that they had no
> objections to publishing them. I wonder who's court the ball is in
> now? I'll go digging.


I found:

http://search.ietf.org/internet-drafts/draft-best-urn-oasis-02.txt

It's not totally clear to me what the URN for a namespace URI would look
like.  The namespace URI is not conceptually identifying a particular
document, rather it's identifying the abstract semantics, which might be
implemented by DTDs, Schemas, stylesheets. What would the "type" field be?
Maybe "ns"?  Since we may have multiple namespace URIs, it seems like the
"document-id" component should not just be a version number. This would lead
to something like:

  urn:oasis:names:tc:relax-ng:ns:structure-0.1

James