On Sat, 2004-06-26 at 15:29, Arkadiusz Miskiewicz wrote:
> > will put
> > Return-path: <my-ml-list-bounces+user=bleble.pl@???>
> > instead of
> > Return-path: <my-ml-list-bounces+user=domain.pl@???>
> > and mailman won't be able to find out for which user the bounces belong.
> Errors-To are even more interesting:
> Errors-To: my-list-commit-bounces+user=bleble.pl+user=bleble.pl@???
>
> Adding information to exim+mailman howto about such potential problems would
> be good.
Why does this sort of thing come up the day *after* I ship a copy to
Cambridge to print up for the upcoming exim course :-)
Could you try the following replacement transport (the only difference
is selective replacement of local_part & domain with original_local_part
& original_domain.
It certainly should fix the problem but I broke my Mailman test rig into
pieces over the weekend so cannot do a decent test on it right now.
Would appreciate it if you could let me know if this works right.
[This *will* wrap... you need to fix that]
# Mailman VERP envelope sender address formatting. This seems not to use
# quoted-printable encoding of the address, but instead just replaces the
# '@' in the recipient address with '='.
#
mailman_verp_smtp:
driver = smtp
# put recipient address into return_path
return_path = \
${local_part:$return_path}+$original_local_part=$original_domain@${domain:$return_path}
# must restrict to one recipient at a time
max_rcpt = 1
# Errors-To: may carry old return_path
headers_remove = Errors-To
headers_add = Errors-To: \
${local_part:$return_path}+$original_local_part=$original_domain@${domain:$return_path}
Nigel.
--
[ Nigel Metheringham Nigel.Metheringham@??? ]
[ - Comments in this message are my own and not ITO opinion/policy - ]
