Mark Mielke [[EMAIL PROTECTED]] wrote:
> The thing is... this makes perfect sense for 'L'ist reply... but not
> for 'g'roup reply. There _is_ a difference.
Yes, there is. 'L'ist reply replies to just the one address that is the
list, while 'g'roup reply replies to all To: and Cc: addresses, though
this can be superseeded by MFT.
> I just don't like this magic happening underneath me. That is why I
> chose to use 'g'roup reply instead of 'L'ist reply.
>
> Now I'm sitting here being accused of having a broken mailer, being forced
> to agree that it is in fact behaving wrong in my situation... and having
> to whine to the makers of the client who don't seem to agree.
What is behaving wrongly? Mutt does 2 basic things regarding MFT:
- for group replying, it honors it if it sees it
- for composing, it *can* automatically generate the header for mail sent
to know lists
If your problem is with the first, it's basically that the people you're
replying to are not meaning what they put in the MFT header. If your
problem is with the second (as it sounds like it is)... Mutt's doing just
what you tell it, and you can turn it off.
> Regardless of how it is implemented now... these are the two functions
> people want:
>
> 1) A method to "Reply to All". No magic except that addresses which
> match the "alternates" variable would be filtered from the To:/Cc:
> lists.
Why no magic? If the sender specified an MFT header, it's proper to think
they meant they really want follow-ups sent to those addresses specified.
If they bothered to include it, why not bother to respect it?
> Here's a few sample cases:
>
> 1) I belong to mutt-users.... some person who doesn't posts a question
> to mutt-users... I want my smart 'L'ist reply function to respond
> to both mutt-users AS WELL AS the person.
If it did that it would not be a list reply, it would be a group reply, by
definition.
> If the person isn't copied then he/she may never get the response!
Indeed. That's why in this case most people say "Cc" me or generate their
own MFT header that looks like the one you mentioned below. If a
discussion starts on a mailing list, the sensible default assumption is
that it should continue there, and the person should specifiy in some way
if otherwise.
> For this case, the desired Mail-FollowUp-To: would include:
>
> [EMAIL PROTECTED],
> [EMAIL PROTECTED]
>
> 2) I belong to perl5-porters, but not perl-loop (I belong to both...
> but pretend I didn't...) Somebody crossposts a thread to
> perl5-porters and perl-loop. I take part in the discussion that
> I see in perl5-porters... Somebody decides to reply to a thought
> I brought up... but they decided to remove perl5-porters because
> it was quite specific to perl-loop... but wait? Where did my
> address go? Because the people who made mutt were thinking that
> my name was redundant my name wasn't put in the header going out.
Mutt's not doing *anything* you aren't telling it to do. This behaviour is
caused by *you* indicating it's redundant by making use of the $followup_to
variable. If you don't want it generating that kind of header, and want to
always be included in group replies, turn it off. The default behaviour
(when no MFT header is there) is for a group reply to go to everyone found
in the Cc: line. The normal reason to change this is to avoid getting
duplicates, hence the $followup_to var. If you want duplicates, don't set
it. If you want a more complex MFT header, just write it.
To quote the manual:
==========
followup_to
Type: boolean
Default: set
Controls whether or not the Mail-Followup-To header field is generated when
sending mail. When set, Mutt will generate this field when you are
replying to a known mailing lists.
The purpose of this field is to prevent you from receiving duplicate copies
of replies to messages which you send by specifying that you will receive a
copy of the message if it is addressed to the mailing list (and thus there
is no need to also include your address in a group reply).
==========
> Point in being... in order for my address to be removed from the
> Mail-FollowUp-To:... it must be determined that all the other addresses
> are either not lists... and the only for sure part here is if the address
> is the same as that of the sender address... and if I belong to every other
> list mentioned.
>
> That is the only behaviour which is practical and to be expected.
It's the exact opposite behaviour of that stated in the manual for that
variable. And judging by the number of people that seem to make use of
$followup_to as it does work, not a lot agree with you about what it should
do.
--
Jeremy Blosser | [EMAIL PROTECTED] | http://jblosser.firinn.org/
-----------------+-------------------------+------------------------------
"If Microsoft can change and compete on quality, I've won." -- L. Torvalds
PGP signature
