On Tue, 09 Feb 2010 21:28:25 +1100, Alex Ghitza <aghitza@gmail.com> wrote:
> Ideally, I'd like to be able to see what's hidden behind the 
> "Non-text part: text/html", and if it could happen inline that would be
> great.  I would also entertain the idea of opening an external browser
> window, if that's the way this has to go.

We actually had some support for inline viewing of HTML message parts
for a while. I reverted the code only because it was slowing down
rendering of non-HTML messages. And I did that with the expectation that
the code would appear again shortly with the obvious optimization to not
do any extra work for non-HTML messages.

I know that Keith is using a little script he wrote so that he can hit
'|' on a message and pipe it to his script. The script then uses a
utility, (munpack?) to extract all the various MIME parts to a temporary
directory and then run a web broswer on that directory. Perhaps Keith
would be so kind as to share that script with the community here.

What I'm doing write now is using 'V' to view the raw message and then
running eval-expression (by typing M-:) and then using:

	(mm-display-parts (mm-dissect-buffer))

That isn't ideal because it leaves a bunch of the raw message around,
and leaves a buffer open with a modified version of a mail file in the
mailstore.

I'd definitely like a correct solution. Emacs definitely has no shortage
of already-existing code to parse and render HTML, (even with inline
images, etc.). We just need to find and hook things up properly.

-Carl