On Thu, 27 Jul 2000, Philip Hazel wrote: B
> > So I really would like to be able to have a custom reason for the
> > failure, as with the prohibition_message option, but this doesn't
> > seem to be possible with a receiver verification.
>
> You need to set up a custom verification failure. If example.com is a
> local domain, try a director like this:
>
> custom_fail_verify:
> driver = aliasfile
> search_type = lsearch
> verify_only
>
> and in the file, put lines like
>
> foo.bar: :fail: Left our employ
>
> If example.com is not a local domain, then you have to fudge things up
> using a router that recognizes the addresses somehow, and forces them to
> become local so that you can then use a director like the above. This is
> messy (maybe one day...). Something like this:
>
> route_to_fail_verify:
> driver = domainlist
> domains = example.com
> local_parts = lsearch;/that/file
> route_list = * localhost byname
> self = local
> verify_only
>
Thanks, that's exactly what i needed.
I assume that I have to use a separate router/director for each sub
domain that I handle.
> > I can't find any prohibition_reason for receiver_verification and my
> > debug sessions shows that prohibition_message is never checked when
> > the verify fails.
>
> The prohibition stuff is only implemented for policy rejections;
> verification doesn't count as such. I know; it's arguable ....
>
Well, since I have to use verification to impose receiver policys
I would argue for verification being a kind of policy rejection. :)
(Or argue for real receiver policys)
Anyhow, Thanks for exim, it sure makes my life a little bit easier.
/Andreas Pettersson
