On Mon, Mar 15, 2004 at 11:37:36AM +0100, Johan Thelmén wrote:
> "the best course of action would be to issue a tempfail after the RCPT
> command. However, in the case of a SMTP callback, doing so at that point may
> cause our outgoing mail to be delayed unnecessarily."
Not to mention that I`m a bit worried that one day some other MTA comes along
(say, somebody suddenly implementing callouts in qmail) and it simply breaks
on deferred callouts.
I think just checking for <> and postmaster (Postfix uses postmaster, not <>
for bounces, AFAIK) and giving them special treatment is a bit ugly; it
smells hackish ;-)
> But the easiest thing is to defer after data.
>
> I'm not against an acl_smtp_predata, it would be good for inserting another
> delay for known spam. But I don't think it should be able to reject.
You mean reject/defer, or just reject? :-) I'd be most happy if I only had
deferring, but of course, the RFCs seem to say even that is disallowed.
/* Steinar */
--
Homepage:
http://www.sesse.net/
