🚀 go-pugleaf

I recently talked to Bill Carpenter about fakemail, which we are going
to use for all our copies of gnuemacs here because they we don't have
to worry if some machine is running sendmail or not.

>>>>> "W" ==  wjc@hos1cad.att.com (Bill Carpenter) e-maild me:

Jacobson> why dont you push to get it adoped by stallman ...
Jacobson> "fakemail will always fail if there is "()" junk in the
Jacobson> address, etc.  so why not fix it?" ...

W> Well, in the first place, he already has my fakemail changes from when
W> I posted them a couple years ago.  I know he looked at them, since he
W> asked me a couple questions about it.

W> In the second place, I don't think there really ought to be a fakemail
W> program.  All that parsing ought to be done in elisp and then the
W> message handed to /bin/[r]mail. [...]
W> I'd whip that elisp version up, but the skill to importance
W> ratio of it doesn't make it past my "do it" threshold.

W> Maybe there is some new way in V19?  Beats me.
W> -- 
W>   Bill Carpenter         att!hos1cad!wjc  or  attmail!bill
W>   (908) 949-8392         AT&T Bell Labs,  HO 1L-410

Anyway, my feelings that I wish to share with bug-gnu-emacs readers is
that, yes, I believe in "what you see is what you sent", meaning that
true, the stripping of reply headers should be done as a reply buffer
is being prepared, (e.g., by VM (has a variable to handle stripping),
or rnewspost.el, (no variable, but I made a hack for it)) but, if the
e-mail is destined to fail in fakemail, then it should be stripped
there instead.  Anyway, like Bill, all this should be done in e-lisp I
say.

Dan_Jacobson@ATT.COM  Naperville IL USA  +1 708 979 6364