Remember, you are stuffing processor requirements in
there (why people dislike PSVI). It is possible that
some boundary exists between well-formed and validated
documents, and so far, no one seems to know just where
that is, yet XML was proposed, designed and sold on the
notion someone would know. What we got instead was a
step by step inching into a system in which semantics
are being inserted via magic strings that are eroding
the boundary between processor and content.
No agreement without a clear and clean requirement.
All you have said is a gap exists. Examples
have been provided to show no gap exists; you don't
like the solutions. That puts the burden of specing a
requirement in the proposer's camp.
Nyet. Not without a clear consensus that
such a thing is really required and that existing
solutions are not adequate. Start there; not with
the end game.
len
-----Original Message-----
From: Daniel Veillard [mailto:veillard(_at_)redhat(_dot_)com]
Hum, not sure I catch your point. Mine is that WG are busy, and
if getting a NOTE in /TR is relatively easy, getting a REC needs
a Working Group scheduled to do it. Doing XML Id would be similar to
do XML Base (retroffitting a change using the gap left in the 1.0
infrastructure) and as much as I would like to see people agreeing
quickly on it and getting it out in a couple of month, I know from
experience it is not possible (again I would love to be proven wrong !)