Re: [exim] XCLIENT supported by exim?

On Wed, Apr 23, 2008 at 08:37:21PM +0200, Marten Lehmann wrote:
> Hello,
>
> > I'm not sure what the benefit is though - surely if you have a device in front of your MTA then that
> > is the place to be doing the rejection or whatever based on where the mail comes from. Surely if you
> > let your MTA do the rejection, you're going to end up with the issue of you generating backscatter
> > spam. Or am I missing something ?
>
> yes. It is not a IronPort or Barracude appliance. It is a realtime
> proxy. No messages are stored. The messages are scanned and passed to
> the backend server in realtime. There is no need to configure lists of
> email addresses to accept. The spam proxy just passes the RCPT TO the
> backend (exim) and if it fails, it rejects the message in the original
> SMTP session. The same applies in case it detects spam. Thats what makes
> the eXpurgate to such a cool and smart solution.

If the message is being scanned by the proxy device, why do you need
to pass the original sender IP to the backend? Presumably, all RBL
checks are being done by the outward facing proxy, yes?

If that is not the case, then it can be argued that the failure here
is in the way your mail system has been implemented.

At the very least, it should bear some consideration that in all the
years and hundreds of thousands of Exim users, the list can only
vaguely recall a few instances of someone asking for anything similar
to what Postfix's XCLIENT hack tries to accomplish. There's a reason
for that.

--
Dean Brooks
dean@???