Re: [XML-URI] HTTP extensions framework comparison

> At 10:08 AM 7/27/00 -0700, Henrik Frystyk Nielsen wrote:
> >Because there is no link between the equality rules and HTTP, there is
no
> >difference between when HTTP wants to compare URIs and other
> >specifications want to compare URIs and it therefore doesn't make sense
to
> >define different rules for comparison when the comparison in fact
doesn't
> >depend on the context (HTTP, or other specs).
>
> As much as I'd like to have a single algorithm for comparing URIs, I
don't
> find the claim that 'there is no link between the equality rules and
HTTP'
> when the comparison rules you're describing are in the HTTP
specification,
> not RFC 2396.

You don't have to take my word for it but this is for historical reasons
rather than anything else. The URI and the HTTP spec were in similar
process state but because the URI spec was rather controversial, it was
not desired to make the dependency. Regardless, the description (which I
would put much more weight on) doesn't indicate such a link and this is
after all what is interesting.

> Unless another document describing URIs _in general_ explicitly makes
the
> claim that the HTTP rules are to apply across URI contexts, I can't find
it
> remotely believable to accept that as the case.  If that should be the
> case, I would strongly recommend revisiting and revising RFC 2396.

I have advocated this on several occasions and I could not agree more. I
think this would be a tremendous help for the whole Web community as it
would have solved many of the problems that led to this mailing list.

> To back to what Larry Masinter said:
> >> There was never any intention that the equivalence rules in RFC 2616
> >> would automatically apply to any other context for URIs.
>
> It really feels like the http scheme is complicating namespace URI
> discussions in an astonishing number of ways.

Not really - it so happens that HTTP URI (as pretty much the only ones)
use all aspects of URIs and have been used in more contexts so that it is
not surprising that this is the case.

Henrik

Received on Saturday, 5 August 2000 01:51:34 UTC