On Wed, 7 Oct 1998, Andromeda wrote:
> At 15:01 07/10/1998 +0100, Philip Hazel wrote:
> >cannot route/direct or to addresses it cannot route/direct. This does
> >not of itself prevent unauthorised relaying, of course.
>
> It should... shouldn't it? Or must I use the receiver_verify and
> sender_verify in a director to check that if the recipient is not one of my
> users, the sender MUST be? And vice versa?
Checking for verifyable senders/receivers does not of itself prevent
relaying. Checking for unwanted relaying is an entirely separate and
independent process. However, they do happen in sequence. If a sender
address fails the verification checks, then there's no scope for relay
checking, since it never proceeds far enough to read the recipients, for
example. There's an amazing flowchart in section 41.6 which shows how
these things interact.
> If so, should I use that other director example you posted earlier?
It sounds as though you need to make use of sender_address_relay and
relay_accept_host_or_sender for your relay controls.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
--
*** Exim information can be found at
http://www.exim.org/ ***
