On Tue, 11 Jun 2019 at 09:12:06 -0700, Kevin J. McCarthy wrote:
> On Tue, Jun 11, 2019 at 12:26:47PM +0200, rear1019 wrote:
> > The ticket assumes that both $reverse_name and $reverse_realname are
> > set. In this configuration mutt should reuse the To: header of the
> > message replied to as-is in the reply’s From: header, even if the
> > name is empty.
>
> The documentation certainly says "as-is", so from that side I support
> you.
Derek Martin provided the crucial hint by explaining the way he
interprets the very last part of:
When [$reverse_realname] is set, mutt will use the address from
incoming messages as-is, possibly including eventual real names.
I have misinterpreted this last subclause. Now I think it means “*if
possible* mutt will *eventually* include the real name”. As Derek
pointed out this is only possible (and sensible) if there is no name in
the message replied to. Only then mutt eventually inserts the real name.
> […]
> If I adjust the documentation, I doubt there will be any complaints.
> But if I adjust the behavior, it is almost certain some users will
> notice and complain vociferously. So, after more thought, I am
> leaning toward adjusting the documentation.
>
> Obtaining the behavior you desire can be accomplished by unsetting
> $realname and instead putting the name inside the various $from /
> my_hdr settings you use.
Some quick testing indeed shows that unsetting $realname and putting it
inside $from has the behavior I want – reusing of To: header of incoming
message as-is (without any modification) as From header in a reply.
$reverse_{name,realname}$ must be set of course.
As far as I am concerned, adjusting the documentation to be more clear
is fine.
