moore(_at_)cs(_dot_)utk(_dot_)edu (Keith Moore) wrote on 22.05.01 in
<200105222234(_dot_)SAA28890(_at_)astro(_dot_)cs(_dot_)utk(_dot_)edu>:
there are lots of ways to describe the difficulty with getting a reply
behavior that satisfies everybody. for one thing, I don't think we've
even come close to getting consensus on what is "proper", much less
getting agreement on a protocol that encourages such behavior.
most of the discussions have been limited to existing mechanisms. that
will never be satisfactory.
well, DRUMS was constrained by charter to only discuss existing mechanisms.
I agree that we won't solve the list reply problem (or any of the other
reply problems) in this way.
a number of extension mechanisms have also been discussed, but most of them
were overly simplistic.
I still like my idea to give a labelled list of possible reply targets,
from which the user could select. Some of these might be standardized,
others might only make sense for a specific message.
That's not a lot different from the mechanism some MUAs already use, which
is to present some sort of list of all the addresses found in the header,
plus one or two default choices (as in, "personal reply" and "reply to
all").
One possibility to do that would be a Reply-Targets: header with a list of
group addresses, like this:
Reply-Targets:
list-submit: ietf-822(_at_)imc(_dot_)org;
list-bot: ietf-822-request(_at_)imc(_dot_)org;
list-owner: owner-ietf-822(_at_)imc(_dot_)org;
conference-papers: ietf-2002-papers(_at_)ietf(_dot_)org;
list-abuse: abuse(_at_)imc(_dot_)org;
rfc-editor: editor(_at_)rfc-editor(_dot_)org;
MfG Kai