On Wed, 18 Oct 2000, Philip Hazel wrote:
> On Wed, 18 Oct 2000, Dr Andrew C Aitchison wrote:
>
> > If the chain reaches a single, non-local address I propose that we accept.
> > (interrogating a remote system may cause hang-ups or timeouts, not to
> > mention loops).
>
> Why? This is no different to verifying a non-local address in the first
> place. Note that Exim does not interrogate remote systems when verifying
> remote addresses[*]. It just interrogates the DNS (assuming standard
> configuration). And actually, it has to do this anyway, in order to
> determine (in some cases) that the address is local.
>
> [*] Some MTAs do do this. I've wondered about it, but it would be
> prohibitively expensive, and not useful against hosts that don't
> verify at RCPT time.
Sorry. I meant not to do [*]. I hadn't realized that we had already
considered this case, and wished to make it explicit.
Sounds as though Phil was going to do what I wanted anyway.
--
Dr. Andrew C. Aitchison Computer Officer, DPMMS, Cambridge
A.C.Aitchison@??? http://www.dpmms.cam.ac.uk/~werdna
