ietf-822
[Top] [All Lists]

Re: [Fwd: I-D ACTION:draft-moore-mail-nr-fields-00.txt]

2004-09-15 08:36:57

Charles Lindsey wrote:

Well there are at least three places it could put information to help a
subsequent MUA:
    Reply-To:
    Mail-Follouwps-To:
    Mail-Copies-To:
and if none of those suffices, then we could always invent another one.

Only one of the infinity of field that you propose is a standard
field recognized by UAs and MTAs.  Any other than that one would
require updating all UAs and many MTAs.

* Should replies to list postings *generally* go only to the list or to
 the list and the user, as a matter of list culture?


Speaking *generally*, the list-owner is in the best position to know the
culture, and that will affect how he fills in whatever header he chooses
to insert.

No, the list owner has no idea of what the message author intended;
he might not be subscribed to the list, desiring personal responses
to be summarized to the list, he might read the list from a web
archive interface w/o being subscribed to the list, he might be
sending a notification to one list that a topic of potential interest
to that list's subscribers is being discussed on another list (with
discussion supposed to take place on the latter list), etc...

But relying on the User to fill in Reply-To would be even worse (indeed
any solution that requires manual action by the user on _every_ response
to a list is doomed from the start).

Your assumptions are unfounded. Many UAs provide for an easily-changed
"identity" that carries a default Reply-To field, or can use a folder-
specific Reply-To default.  There is no *requirement* for manual action;
moreover with UAs that are likely to be heavily used, the defaults set
up (one time) will be appropriate.

In particular, use of Reply-To is the
Wrong Thing if the replier really does want to reply to the author and ot
to the list.

No, a respondent could (and Keith would say "should") use the From
field, not Reply-To, if he specifically wants to respond directly
to the author(s) only.

OTOH it does work quite well when inserted by the
list-expander in particular lists with a culture suited to it.

Only -- as we seem to agree -- if an author-supplied field is
never overwritten.

Mail-Copies-To is all right so far as it goes, but it only tackles some of
the problems. I think, on balance, I would prefer Mail-Followups-To.

Neither is standardized or recognized by a substantial fraction
of UAs or MTAs.




<Prev in Thread] Current Thread [Next in Thread>