Re: lack of consensus on httpRange-14

Tim Bray writes:
> Well, the key thing is that we couldn't find many implications. 

I'll readily concede that the implications are not technically obvious.
URIs work just fine provided that:

   a) you don't have to think about what the URI means
   b) you don't worry about whether the URI has (the same|any) meaning 
       in other contexts

(a) is the pitfall that's generated thousands of messages on XML
namespaces, while (b) is the pitfall that seems to arise whenever specs
use URIs in ways that go beyond ordinary (vulgar?) URLs.  Developers who
are used to transparency with URLs (404 as worst likely result) can be
boggled with URNs, and aren't clear on what adding a magic # sign means.

> If
> we had then it'd probably be inappropriate to de-prioritize it.  So
> if someone comes forward with concrete examples of impact, we're
> probably gonna have to go back to work.  Anyhow, we didn't say
> consensus was impossible, we just said that
> 
> (a) we don't have consensus at the moment
> (b) we aren't convincd that the issue is impactful enough to invest 
> the  work

Given the nature of the discussion and the objections, I'd strongly
encourage the TAG to conclude that consensus is impossible.  This isn't
a new discussion - these conversations have been going for years.
 
> By the way, TimBL may well be on your side on this one; he's got
> worries that this one will bite the Semantic Web work.

I suspect from prior correspondence that TimBL would prefer to see URIs
sorted out generically in a manner which conforms to his expectations
about the Semantic Web rather than just "there is no solution, so RDF
can use # if it wants."  

I'd be interested to hear if his perspective on the matter would support
different URI usages for different applications.  

> I would prefer to keep this issue alive because on the face of it's
> irritating that we don't have a clean clear answer for "what can a
> HTTP URI identify?" 

>From my perspective, it's better simply to say this is irritating than
to try to create a one-interpretation-fits-all interpretation of URIs,
even HTTP URIs.  Given the tangles with fragment identifiers and
representation types, there may not be such an aswer.

> or more concretely "is it OK to use
> http://www.w3.org/ when I'm making RDF assertions about industry
> consortia?"  -Tim

My suggestion would let the RDF specification make the concrete decision
about whether that (or http://www.w3.org/#) is okay or not, without
worrying about its impact on other specs.

-------------
Simon St.Laurent - SSL is my TLA
http://simonstl.com may be my URI
http://monasticxml.org may be my ascetic URI
urn:oid:1.3.6.1.4.1.6320 is another possibility altogether

Received on Thursday, 3 October 2002 14:08:23 UTC