- From: Larry Masinter <LMM@acm.org>
- Date: Thu, 29 Nov 2001 09:21:03 -0800
- To: <Patrick.Stickler@nokia.com>, <fielding@eBuilt.com>, <ejw@cse.ucsc.edu>
- Cc: <w3c-dist-auth@w3.org>, <uri@w3.org>
> > In other words, I think that "scheme:" is only a valid
> > identifier for the
> > namespace if the scheme defines it as such.
>
> Fair enough. Though it seems that an RFC revision would
> still be in order to permit schemes to define "scheme:"
> as a valid absolute URI -- and also once some schemes
> adopt such a practice, it will be pretty darn hard to
> keep folks from presuming that "foo:" is the canonical,
> official URI denoting the scheme.
It's pretty darn hard to keep folks from presuming that the
moon is made from cheese, too, despite the fact that it
wasn't when we got there. So if the official RFC says
that an undecorated "scheme:" is only valid if the scheme
definition says it is, and it only means what the scheme
definition says it means, well, we'll just have to put up
with those cheesy jokers who decide it means something else.
> It may be more practical to just state it once and
> for all in a revision of the RFC. Otherwise, all
> existing schemes that would like to use "scheme:" would
> themselves have to be revised.
How does a URI scheme 'want' to do something?
If YOU want to identify the URI scheme itself, you need
something else. I would suggest using the process in
http://www.ietf.org/draft/draft-mealling-iana-urn-02.txt
making a "uri-scheme" space, and then using, say,
urn:ietf:uri-scheme:http
if you want to talk about the "http" URI scheme.
I think it must be the case that "http:" remains illegal
as a URI even though "DAV:" would be allowed.
Received on Thursday, 29 November 2001 12:24:20 UTC