AW: Are *relative* URIs as namespace nemes considered harmful?

I'm risking to ridicule myself among people I value highly here,
but please excuse my ignorance and allow me to ask a few
questions:

	When are 2 URIs equivalent?

I've seen many people assuming that we know and that it's only a
question whether to absolutize or not, but looking up RFC2396, I
find that "equivalence" depends on the schema used (section 6)
and rules for absolutizing URIs independent of their schema
(section 5). Building on the assumption that namespace URIs in
fact identify "bags of element names", I wonder

	how are we supposed to do identification if we can't establish
equality?

If this is true: Wouldn't it help to fill that gap and specify
when URIs are equivalent at least for the most common schemas
(schemata :-)?

Then, we can deprecate current practice as the necessary fix it
was in a situation without that specification.

Josef

P.S.: "current practice" in this context is comparison of URIs in
various ways as done in several XML specifications.
--
Stack Overflow - We Help Increase Your Productivity With XHTML
Visit http://www.mozquito.org

Received on Monday, 22 May 2000 11:58:58 UTC