RE: [namespaceDocument-8] 14 Theses, take 2

On Tue, 2002-02-19 at 06:33, Graham Klyne wrote:
> At 01:20 AM 2/19/02 -0500, Simon St.Laurent wrote:
> >I'm not sure my being convinced or not matters, so long as you're in the
> >business of making sure there's a level of indirection between namespace
> >URI and schema XYZ.  Lose that level, and I'll argue quite strongly.
> 
> Noting some points from DanC's argument, I'd soften this to allowing that 
> indirection is possible, but not always required.  Then, for cases where 
> simple content negotiation is sufficient, the indirection can be avoided.

And that will make me argue quite strongly.  Take away the requirement
of indirection, and you won't have it - the primary argument for
slapping schemas at the end of URIs appears to be laziness. (Not that
laziness is always a bad thing, but it often has hidden costs.) 

I don't find content negotiation useful in this case, if only for its
lack of transparency.

Try resolving:
http://www.w3.org/2001/XMLSchema

in both Mozilla and Microsoft Internet Explorer, and you can marvel at
the difference.  I could have found the schema from Mozilla - as could a
program - using the RDDL file sent to Internet Explorer, but instead
it's just dropped on my head.

Yecch.
 
-- 
Simon St.Laurent
Ring around the content, a pocket full of brackets
Errors, errors, all fall down!
http://simonstl.com

Received on Tuesday, 19 February 2002 09:47:00 UTC