Re: Closure on canonicalization, I hope

I was just on the verge of finishing up a paper explaining the issues,
and proposing use of UTF etc. I will still send this out as it's
relevant for browser writers, and HTML authors.

Just to clarify, in Larry's proposal, the Accept-charset is not required
because we can do the same thing via:

   Accept: text/html; charset=xxxx

However, how do we handle text/sgml, text/plaintext etc? Can they be
handled the same way, or do we still need Accept-charset.

I will be sending out my text tommorrow.