I'm waiting to see Gavin's proposals on this. I'm assuming
he's going to do something with ERCS.
I took a look at the ERCS (Extended Reference Concrete Syntax) page
he posted, and it looks promising as a way to write SGML declarations
for larger character sets. On the other hand it looks like
ERCS is itself in draft status. It might make sense, if the
working group agrees with the concept of using ERCS to support
HTML over Unicode (and other character sets sufficently unlike
Latin-1 that are subsets of Unicode), to try and put hooks for
it in the main standard but write it up as another internet-draft/RFC.
We've already got some of this, I'd guess a big issue is how the
group feels about different SGML stuff for other character sets.
(If people really don't like ERCS, we could take the least-common-denominator
approach of forcing all markup to be in the same bunch of characters
as Latin-1; so no ERCS != no Unicode.)
Has anyone actually looked at what Netscape is doing with
character set support in their new version (I'm not in
a position to judge the results.)
The other thing I was pushing for was a language tag/attribute; but
that's coming in HTML 3.0.
I'd suggest that a good critera for trying to put something
in 2.1 is that it be modular and not drag in lots of side
issues.
-- Albert Lunde Albert-Lunde@nwu.edu