David DeSimone <[EMAIL PROTECTED]> wrote on Mon, 27 Sep 1999:
> For instance, perhaps I am reading mail in my mutt-users folder. Then
> something jogs my memory, and I realize I need to send mail to my friend
> Joe. Are you saying that, just because I am viewing mail in my
> mutt-users folder, that my letter to Joe should be saved there instead
> of my usual Fcc?
No, I'm not saying that. The question was about save-hooks, not
fcc-hooks.
Then again, I can see there being a need to save a FCC to a different
folder than the default when sending mail related to mails in a specific
folder, but that can probably be accomplished with properly crafted
send-hooks. It would be quite impossible for mutt to determine whether
mail sent while viewing a specific folder was "related" to that folder
or not.
> You see, it shouldn't matter what folder you're in. You should be able
> to construct a send-hook that triggers on real information, such as whom
> the letter is two, or from, or Cc'd to.
Correct, for send-hooks.
> If you're in your mutt-users folder, and the
> mail that you now send is related to that mailing list, then that
> address probably appears in the headers, and so you should be able to
> construct an fcc-hook or save-hook that does what you want.
Ahh, but I can see more complex issues for constructing save-hooks
than that. You see, procmail and other mail filtering tools allow for
a very complex methods of filtering. Mostly these are header-based,
but they can also be based on filtering programs or other rules based
on program output. Therefore it's entirely possible that Mutt can't
reproduce the rule why a given message was sorted to a particular
folder. And so you can't have a simple save-hook that could follow
the same rule. Thus, you need folder-specific save-hooks for messages
in that folder. Currently the only way to do that is with setting a
save-hook from a folder-hook, but that has the drawback that you can't
undo it.
Admittedly this is a rare case, so it can be argued that the need for
such functionality is not worth adding.
Mutt doesn't allow for "undoing" of hooks. That's fine, I can
appreciate that it would be too complex to try to keep record of
different states. You can only "re-do" hooks. The problem here is
that the default behaviour for mutt for the save command (using the
user's username from the email address) can not be re-established in
any way once you have specified a default save-hook (save-hook .
+folder, or using ~A, or whatever). The only way to avoid that is
never to establish a default save hook, but this limits you so that
you don't have a complete freedom to define save-hooks which behave
exactly like you'd want.
Maybe then the desired solution would be to have a "magic save-hook"
that re-establishes the default behaviour? Or is such a thing already
possible, if "save-hook ~l =%B" works? All that would be required is
to have a %-variable which could be substituted for the username, and
it would be possible to use "save-hook . =%u" to restore the default
behaviour (using %u in this example).
Mikko
--
// Mikko Hänninen, aka. Wizzu // [EMAIL PROTECTED] // http://www.iki.fi/wiz/
// The Corrs list maintainer // net.freak // DALnet IRC operator /
// Interests: roleplaying, Linux, the Net, fantasy & scifi, the Corrs /
Today is the last day of your life so far.
