On Mon, 30 Sep 2002 01:29:25 +0200
Thorsten Maerz <torte@xxxxxxxxxx> wrote:

> The reply function shouldnt even be changed by a default-reply-to
> option. If reply-to-list can be improved to detect any list (with
> a unique header) and overrides reply-to:, the patch can be reverted.

I was thinking about this one; I'm glad you brought it up.  I think that
something like this is needed to deal with lists that insist on munging
headers.  In such situations, reply to list will already do the expected
behavior, so in order for reply to sender to work, the from header, not
the reply to header (which has been munged) would need to be followed. 
Maybe there would need to be a configuration option about this one. 
Anyway, if this wasn't done, we'd end up with 2/3 functionality again,
as two of the reply functions duplicate each other.