- From: Martin Duerst <duerst@w3.org>
- Date: Wed, 05 Feb 2003 12:40:48 -0500
- To: "Larry Masinter" <LMM@acm.org>, <chris@w3.org>, "Tim Bray" <tbray@textuality.com>
- Cc: <xml-names-editor@w3.org>, Michel Suignard <michelsu@microsoft.com>
Hello Larry, Many thanks for your mail. When I looked at http://www.w3.org/TR/xml-names11/#IRIComparison over the last days, I was thinking exactly in the same direction. At 07:58 03/02/05 -0800, Larry Masinter wrote: >I think there's quite a bit of discomfort with >section 2.3 because of its implication for IRI >deployment: > > http://www.example.org/rosé [I have substituted é for the actual character because it doesn't survive my Japanese mailer :-(] > http://www.example.org/ros%c3%a9 > http://www.example.org/ros%c3%A9 > http://www.example.org/ros%C3%a9 > http://www.example.org/ros%C3%A9 > > >I think the problem is that the Candidate Recommendation >makes no recommendation about the possible simultaneous >use of all of these as namespace names. Yes indeed. It should very clearly say something about this topic. >I think the simplest way of resolving the possible >contradiction would be disallow hex-escaping in >IRIs used for namespace names. That would be one possible solution, and a very clear one. There are a few details to be worked out. For example, do we want to disallow all %-escapes? Or %-escapes above %7F? Are there any legacy namespaces that we have to care about? >More generally, it would be a good idea to strongly >discourage the simultaneous use of multiple URIs >or IRIs that might, for some applications, be considered >equivalent: > > http://www.example.org and http://www.example.org/ > >should not both be used as namespace names even >though they are not string-equal. I fully agree with this, and think that this should be fixed when going to Proposed Recommendation with http://www.w3.org/TR/xml-names11. Regards, Martin.
Received on Wednesday, 5 February 2003 12:45:59 UTC