On Mon, 19 Jul 1999, Dave C. wrote:
> > > > headers_add = "X-POP_authentication: ${lookup {$sender_host_address} \
> > > > lsearch {/etc/exim/pop_smtp_auth.users} {$value}}
> >
> > What if I also used net32-lsearch here? I think I'll go have a try at
> > that...
You don't need anything fancy here. You are specifing the key yourself.
If you want to use a net32-type file, you could always write
.... ${lookup {$sender_host_address/32} ....
> Ugh. It didnt recongnize this here. Also, for the host_accept_relay
> check, when doing a -d9 with sender_host = 10.1.2.3 (contrived
> example), it was looking for "3.2.1.10/32" as the lsearch key? Is
> something reversing the octets for this? Is that right?
Oh my goodness. It shouldn't be doing that. What hardware architecture?
It is my believe that the existence of both "little endian" and "big
endian" processors has caused more bugs of this type than most other
things.
I think that is probably what's caused this. I'll investigate.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
