RE: Irony heaped on irony

> The main aspect of it that I see as rushing is this quiet shift to using
> namespace URIs to point to schemas.  This controversial practice is being
> put forward through the schemas group, for which its useful, with no
> apparent discussion of its impact on other XML activities inside and
> outside the W3C.

The Schema WG is not advocating that schemas be pointed to by namespace
URIs. That is what the SchemaLocation attribute is for. Do not confuse the
opinions of some as the consensus of the Schema WG.

> This activity is a large part of what makes the relative URI discussion so
> ugly, because the results of that are intertwined with the
> NSURI-pointing-to-schemas issue.

That is a red herring IMHO. The issue is that Namespace, XPath, and DOM recs
are inconsistent with their use of NS URI equivalence. There is currently no
rec stating what an application can or can't do with a namespace URI.

Decide what rec needs to be changed and do it is such a way that doesn't
break currently conforming documents. Leave the packaging and dereferencing
out of it for now.

> I think the experience has already exposed problems, as the W3C
> has started
> putting schemas at the locations referenced by their namespace URIs and
> received some loud comments from people that this may not be so good an
> idea.  Is this experience part of testing XML Schemas?  I'd argue that it
> shouldn't be, that it in fact reflects a change in practice more relevant
> to namespaces than schemas.

Since there is currently no specification as to what resides at the endpoint
of a NS URI, shouldn't it be up to the owner of that URI to decide what is
there? The fact is that in many real applications, schemas will not
retrieved over the web to process an instance document. They will be stored
and retrieved locally based on the NS URI. But again, this has nothing to do
with solving the problem at hand.

David Cleary
Progress Software

Received on Friday, 19 May 2000 11:54:46 UTC