In <20050526041541(_dot_)GA12769~(_at_)nicemice(_dot_)net> "Adam M. Costello"
<ietf-822(_dot_)amc+0(_at_)nicemice(_dot_)net(_dot_)RemoveThisWord> writes:
There are several things that would be useful to be able to express.
Here are the ones that spring to my mind right now:
* What sorts of replies are allowed/encouraged:
- Prefer private reply.
- Prefer public reply.
- Public reply is the only option (From address is bogus).
- No preference.
(Mailing lists may want to supply a preference if the author did
not supply one, but they generally should not clobber the author's
preference if one was supplied.)
* Where private replies should go:
- The From address vs. some other address(es).
* Where public replies should go:
- To the To/Cc addresses, or to some other addresses? (For example,
replies to an announcement list might be directed to a discussion
list.)
- Should the private-reply address be excluded to avoid duplication
(because the public-reply addresses already cover it), or included
(because the public-reply addresses don't already cover it)?
I think we established last time round that the MFT header could cope with
all those requirements, provided the standard was clear as to how it was
to be used, and who should be able to write (or not overwrite) it.
Main problems with MFT are
1. It presumes that MUAs will have at least three buttons corresponding to
reply (to author), Reply-to-All, and Reply-to-List/Followup/whatever. You
can't mandate in a standard that all MUAs should provide all three, but
you can indicate that the standard assumes them (and then specifies who
the mail should be sent back to - modulo user overrides - in the various
cases). There also needs to be a fallback interpretation for MUAs that
only provide the first two (as is commonly the case at the present time).
2. It really needs a sympathetic MUA which can be configured by the
original author to set the MFT header as he desires. I gather that MUAs
written by DJB do this nicely, but sadly there exist MUAs not written by
DJB :-( .
3. It needs to be defined how it interacts with the various List-*
headers. Sadly, many mailing lists still do not use those headers.
4. It really needs to cope with News as well, in place of the
Mail-Copies-To header that does not quite do the right thing for mail (and
which the Usefor WG has decided not to standardize at present, partly
because a solution that works for both Lists and News would be better).
--
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: +44 161 436 6131 Fax: +44 161 436 6133 Web: http://www.cs.man.ac.uk/~chl
Email: chl(_at_)clerew(_dot_)man(_dot_)ac(_dot_)uk Snail: 5 Clerewood Ave,
CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9 Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5