On Thu, Oct 15, 1998 at 09:45:26AM +0100, Philip Hazel wrote:
> On Wed, 14 Oct 1998, Ilya Ketris wrote:
>
> > I wonder if there is any possibility of
> > hooking a filter on the processing of the
> > incoming SMTP messages, in the way it's
> > done by the transport_filter option.
>
> Suggestion noted, but..
>
> > I would prefer to say a spammer or a relay site
> > "no" in the way they'd see this, rather then
> > silently disposing the message after telling
> > the sender that it was "250 OK".
>
> The result would be to send 5xx after receiving the data (that is, it
> would be received by the sender after sending the terminating "."). It
> is well known that many MTAs do not treat this as a permanent error.
> They just try to send the message again later, for days and days. So the
> facility might well not achieve what you want.
It will achieve exactly what I want -- first,
the message will never reach my server, and
second, it will make the message stuck on the
spam server and punish it by lying around for
days and days which in turn, may draw the
admin's attention to the problem.
Once a piece of spam is received by my host,
it's considered sucessfully delivered and the
sender couldn't care less if I read it, dispose
it, or my MTA or MUA filters it for me -- their
goal is achieved, and harm is done.
But if everybody would refuse even accept this
piece of spam, the fact of relaying would very
quickly become evident to administrators of
the involved mail relays, they would most likely
react promptly to protect their systems.
Besides, we can detect repetetive attempts to send
a message that was denied in such fashion and
will give it 5xx on the next try after RCPT TO:.
--ilya
--
*** Exim information can be found at
http://www.exim.org/ ***
