Charles Hines writes:
So, I guess I'm looking for a couple of hints on how to either
get it to always display the text/plain portion instead of the
text/html portion, or how to get around that "bad" character
(if I remember correctly it's some encoding of a single-quote,
at least that's what it was on past messages I saw with this
problem). Is anyone else seeing this problem and figured out
how to work around it? Since this particular email went to the
list, I figured that if anyone else was seeing this problem now
would be the time to ask. :)
If you build XEmacs with MULE disabled I think the problem will
go away.
My current working theory is that when VM calls decode-coding-region
on a region that contain one of those strange Windows-only characters
like \222 it gets converted into something that's either invalid as a
character or something unexpected by code further along. If you
could provide a Lisp backtrace, we could determine whether it is VM
or W3 that is croaking, and proceed accordingly.
(require 'vm)
(load "vm-mime.el")
(setq debug-on-signal t)
and then reproduce the problem to get the backtrace.
I'm thinking that I should have VM enforce the iso-8859-1 encoding
space by stripping the high bit off characters from 0200-0240.