Anne van Kesteren wrote:
On Tue, 24 Feb 2009 16:46:59 +0900, Rob Sayre <[email protected]> wrote:
It seems to me that the requirements placed on documents are not
required for interoperation.
Surely for interoperation with validators, editors that refuse to load
documents that do not meet the requirements, and future software that
might do different things with invalid documents it is important that
documents meet the requirements?
I don't believe that interoperation with validators (or test suites) is
what RFC 2119 is meant to address. I have plenty of friends at the IETF
who I am sure could help address this part of the question.
Given the statistics I have seen thrown around, an editor that refused
to load documents which aren't conformant would find little use. Do you
know of such editors?
Future proofing is a valid concern, but I don't think it applies to the
specific cases that we have been discussing recently. Marking summary,
profile, or even property non-conformant will have very little near-term
benefit in terms of making such attributes available for future
revisions of HTML.
- Sam Ruby
