Re: mail-followup-to / mail-copies-to

On Wed, 25 May 2005, Paul Jakma wrote:

> I'm sceptical the Reply-To approach would work - I think i'll try an 
> experiment and see how it works in practice.

Early results suggest Reply-To is not a good idea: Gnus and 
Thunderbird use Reply-To always (Reply and Reply-all).

The classic "Reply-To considered harmful" argument on list mail 
applies to this usage of Reply-To surely?

(Pine is slightly better, it tends to ask whether the Reply-To should 
be followed).

There's also the problem that the Reply-To isn't transitive. Any 
reply will (likely) be:

From: respondent
To: sender, list

Anyone who replies to /this/ mail on the list likely will reply to 
just respondent and list - indeed, it could just be to the list 
(there's seems to be list netiquette developing of stripping out all 
addresses bar the list. Or MUAs like Mutt use the List-* headers to 
only include list if the "list reply" functionality is used, and also 
copying Mail-Followup-To listed mailboxes - hence why I list that 
problematic header in my mail).

I wonder, if one considers:

- Reply-To is one shot, not transitive through a thread

- Reply-To makes it difficult for people to /not/ reply to the
   Reply-To. (I know the position on this, MUAs should be
   fixed, but..)

- MUAs exist (Mutt, Apple Mail) which offer a "list reply" function,
   some of which use ad-hoc and incompatible ways to allow originators
   to specify their preference for replies to list mail

That it might be time to consider a list-reply-preference type 
header? The "no choice / always Reply-To without question" behaviour 
of many MUAs aspect particularly worries me and makes me wary of 
using it to indicate list-reply preference.. (i've seen what 
happens).

regards,
-- 
Paul Jakma	[email protected]	[email protected]	Key ID: 64A2FF6A
Fortune:
There are no emotional victims, only volunteers.