David Woodhouse wrote:
And could it also be that your CBV is using a non-empty MAIL FROM:
address, so it doesn't look like a bounce, and is getting rejected
accordingly?
--"Jeremy T. Bouse" <jeremy+spf(_at_)undergrid(_dot_)net> wrote:
Actually at this time I have no CBV in place on my servers. In
this instance it came as a result of me trying to post to a
SourceForge.net mailing list and their CBV tried to verify the
return-path which was SRS signed address.
There's actually quite a few people doing buggy CBV with non-empty
reverse-path. Most of them seem to use postmaster@, so I've made my
implementation of SRS accept mail from postmaster@ to the SRS0 addresses
too.
You might be able to see more info in your mailserver logs, around the time
the message was refused... see if sourceforge tried to connect back to you
and what their CBV uses as the mail-from address.
Perhaps it's a good idea to accept mail to SRS addresses claiming to be
from postmaster@ ... In addition to CBV reasons, I have found a small
number of normal bounces that use MAIL FROM: postmaster(_at_)* or MAIL FROM:
mailer-daemon(_at_)* instead of mail-from: <>
--
Greg Connor <gconnor(_at_)nekodojo(_dot_)org>