At 13:33 +0100 2003/01/20, Tony Earnshaw wrote:
>man, 2003-01-20 kl. 08:54 skrev Giuliano Gavazzi:
>I'm not you know :-) I connect by dialin, so my machine isn't always "on
>the air". But if you let me know, I'll bring it up and you can try to
>relay through it. You'll get a non-standard and unnecessarily rude
>message back.
>
>> If you instead just meant that you are accepting on the basis of
>
> > accept hosts = +relay_from_hosts
>>
>> where relay_from_hosts includes 127.0.0.1, then "my" rule will still
>> work. It just needs to be put right before mail is *accepted* for
>> local users.
>
>I have that acl, yes, but also accept with an acl diametrically opposite
>to yours (accept instead of deny, with an endpass). Yours negates it,
>actually quite funny to see working.
I trust you, but I would still like to have a go (I actually already
tried but, as you say, your machine is not always online).
From what you say you must have something like, I *guess*:
accept sender_domains = +local_domains
endpass
hosts = +relay_from_hosts
message = we do not relay
that, if it is the only "outgoing" accept rule, would effectively
enforce that the MAIL FROM: is in the local domain.
>Philip's comment that the idea isn't good, anyway, holds the sway.
this is also what I said:
At 14:05 +0000 2003/01/19, Giuliano Gavazzi wrote:
>Also, this will stop any emails a local user sent to a remote account
>that in turn get redirected to another local user.
but I do not know what the "roving users" Philip mentioned are.
So, could you dial out at 14:00 GMT (unless your rule is similar to
the one I wrote above)? Is the IP the same as the outgoing one, by
the way?
Thanks
Giuliano
--
H U M P H
|| |||
software
Java & C++ Server/Client/Human Interface applications on MacOS - MacOS X
http://www.humph.com/
