Re: URI resolver was Re: RDDL and XML Schemas Proposed Recommendation

[Michael Mealling <michael@bailey.dscga.com>]

On Sun, Mar 25, 2001 at 03:05:43PM -0800, Tim Bray wrote:
> At 05:18 PM 25/03/01 -0500, Michael Mealling wrote:
> >While it may be safely lost on people (users?), if its lost on system 
> >designers you get huge interoperability problems....
> 
> An example would be helpful to avoid us designers better understand
> the kind of mistake to avoid.

Its mostly in terms of standards. There are a lot of demands for
URIs to be 'fixed' so that you can tell if "two URIs identify the
same resource". This is one of the main questions that the
W3C URI Planning Group was tasked with answering. It came up
in the Future of URIs BOF at the IETF last week as a burning demand
from several participants (they didn't identify what effort had
the perceived problem).


> >Yes it does sound academic but its this exact subtlety that caused
> >so much consumed bandwidth over the XML Namespaces issue. It is
> >very tempting to say things like "I should be able to tell if
> >two URIs identify the same resource" but what you find is that
> >if you don't pay attention to the subtleties in that assertion you
> >end up with a system that breaks a large number of applications very
> >quickly.
> 
> I don't think anyone in the namespace debate made any claims
> quite that oversimplified... once again, an example of the
> kind of errors you see happening (or having a high likelihood
> of happening) would be real helpful.

Some did. There was a rather long thread that the problem wasn't
with XML Namespaces but rather with URIs themselves. I could go
track it down....

> >It all goes back to that question: What does a URI identify and 
> >is that relationship 1:1? 
> 
> ... and is the question about 1:1 meaningful :)?  I don't think
> I'm really disagreeing with what you're saying, just whining 
> that the argument is on too theoretical a level. -Tim

I keep getting asked the question both in the W3C and the IETF
so its at least meaningful enough to those who are asking. ;-)

-MM

-- 
--------------------------------------------------------------------------------
Michael Mealling	|      Vote Libertarian!       | www.rwhois.net/michael
Sr. Research Engineer   |   www.ga.lp.org/gwinnett     | ICQ#:         14198821
Network Solutions	|          www.lp.org          |  michaelm@netsol.com