A draft of the first technical corrigendum to the HyTime
standard has recently been made available. There is a new
Annex defining "formal system identifiers", which, among
other things, provides a framework for handling different
character encodings [*].
It's in Erik Naggum's SGML repository at UIO:
<URL:gopher://ftp.ifi.uio.no/00/pub/SGML/httc1/hi1anfsi.txt>
It looks very similar to the FSI scheme that SP uses.
Another new annex, in the same directory, defines a general-
purpose mechanism for defining architectural forms. This
also looks pretty interesting. It's worth looking at if
HTML 4 is going to go that route.
<URL:gopher://ftp.ifi.uio.no/00/pub/SGML/httc1/hi1anarc.txt>
--jenglish@crl.com
[*] I'm pretty sure that's the right term; UTF8, Shift-JIS,
that sort of thing.