Re: Charset compromise (Was Re: Character-set) header

Greg has an excellent point.  If we are really going to ban nested
encodings, particular encoded message types, this functionality goes
away.  A message/headers subtype would be a workable, if ugly, solution.
 Any others?

(Aside:  There have been about a dozen different proposals, so far, for
how to handle the non-ASCII-headers problem.  None of them, in my
opinion, have been even remotely "pretty."  I, for one, would be
impressed if we come up with something workable, however ugly it may
seem.)