On Tue, 1 Oct 2002, Tom Marazita wrote:
> Philip Hazel wrote:
> > On Tue, 1 Oct 2002, Nico Erfurth wrote:
> >
> >
> >>So you want to accept mails from hosts that would not accept your bounces?
> >
> >
> > Exactly.
> >
> > The whole point of callout is to check that the host will accept a
> > bounce from you before you accept the message. If you aren't going to
> > enforce that, there's no point in doing the callout at all.
> >
>
> I agree; however, it seems the bounce can fail for two (maybe more) reasons:
>
I brought this up once before -
It comes down to wanting to be able to reject outright spam, while not
being in a position to be able to reject legitimate mail from people
whose mailservers are improperly configured. And as much as it is not a
good excuse, the response would still be "Everyone else accepts my mail"
> 1) The RCPT TO: can fail, in which case I'm happy to reject
> the message since it means the sender was invalid. I see a large
> number of these situations in the log files, and they are sender addresses
> that have usually been forged and for which I do not want to accept the mail.
>
> 2) The MAIL FROM:<> can itself fail, which indicates the host at the
> other end has a poorly configured mailer; or at least
> that's what it seems to have indicated with the hosts I've
> come up against recently. The sender might still
> actually be valid though, and in that case I might still want
> to accept the message even if I can't verify it due to their not
> accepting the bounce.
>
> I've come up against about half dozen of these in the past
> six months; in all cases the sender would have been a valid
> address but we rejected the message. Naturally I have contacted
> their postmasters, but with very little response.
>
> My only other options are to disable sender/callout (but it has been
> so useful otherwise...) or keep a manual list of domain exceptions (which
> I have to deal with only once the problem has been identified).
>
> Thanks,
> Tom
>
>
>
>
> --
>
> ## List details at http://www.exim.org/mailman/listinfo/exim-users Exim details at http://www.exim.org/ ##
>
>
