Jay D. Dyson writes:
> I think the folks at vix.com and abuse.net are *far* from
> "idiots." I'm also currently attempting to acquire the source for the
> new-rlytest.cgi script to perform modifications that will demonstrate that
> Qmail doesn't relay. Even so, I think it would be good for Qmail to
> outright reject such relay hacks. I'd rather have a 553 than a load of
> internal bounces in my logfiles.
The problem is not relay checking per se, but the real problem is that
qmail-smtpd does not check whether the local address is valid, before
accepting the message. The relay check that gets accepted looks like an
address in the local domain, so the message is accepted. Only afterwards
does Qmail figure out that the local address doesn't exist, and the mail is
bounced.
--
Sam
