Re: [Fwd: Re: Spamassasin over RBL, was Re: rblsmtpd -t?]

["Jason Lim" <maillist@jasonlim.com>]

> On Tue, 7 May 2002 19:48, Jason Lim wrote:
> > And I also believe in it, because it is very possible that one of the
> > tests to determine if it is an open-relay is braindead... what if I
made a
> > mail server that pretends it will relay email, but in fact does not,
and
> > actually records the IP that tried to abuse the open relay and reports
it
> > to the admins (i consider that very whitehat)? My point is that the
test
> > is not foolproof either... unlike your "everything is black and white"
> > stance.
>
> If you can send a cryptographically signed message is to a mail server
> outside your network and addressed to a machine in your network, if you
> receive it at it's destination and the crypto sign matches then you know
it's
> an open relay.

Note that I said "one of the tests". There are certainly a whole number of
ways to confirm that a mail server really is relaying mail. But I know
some RBLs only employ the most rudimentary of tests.

Naturally, the way you suggest above will provide absolute truth. But do
all the RBLs employ this method? That was the only point I'm trying to
bring up. I was trying to point out that even RBLs that only block open
relays (as opposed to blocking spamming ISP, etc.) can be wrong sometimes
if they do not employ the correct methods, such as the one you mentioned
above.

>
> > This word is not "black or white"... if only it were.
>
> Open relay tests are very "black or white".

...but only if performed correctly.


-- 
To UNSUBSCRIBE, email to debian-isp-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org