On Mon, 15 Nov 2004 10:12:07 -0700, Antone Roundy <[EMAIL PROTECTED]> wrote:
I noticed this Pace doesn't mention canonicalization, and from what I can tell, RFC2396bis section 3 leaves room for variety in how a URI may be formed. Shall we add language similar to that found in section 3.6 (identity constructs), specifically 3.5.1 (recommended normalization strategry) and 3.6.2 (character-by-character comparison)? We could also simply refer to those sections and say "follow the same rules".
I think we should try to define atom:link/@rel as a type of �Identity Construct�. That saves a lot of language, at least.
-- Asbj�rn Ulsberg -=|=- http://virtuelvis.com/quark/ �He's a loathsome offensive brute, yet I can't look away�
