Received: from localhost  
        by CS.UTK.EDU with SMTP (cf v2.9s-UTK)
	id SAA02103; Mon, 5 Apr 1999 18:39:01 -0400 (EDT)
Received: by cs.cs.utk.edu (bulk_mailer v1.12); Mon, 5 Apr 1999 18:38:45 -0400
Received:  
        by CS.UTK.EDU (cf v2.9s-UTK)
	id SAA02045; Mon, 5 Apr 1999 18:38:44 -0400 (EDT)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.12.23]) 
        by CS.UTK.EDU with SMTP (cf v2.9s-UTK)
	id SAA01982; Mon, 5 Apr 1999 18:38:35 -0400 (EDT)
Received: (qmail 8426 invoked by uid 50); 5 Apr 1999 22:38:28 -0000
To: drums@cs.utk.edu
Subject: Re: cs.utk.edu and ORBS
References: <John C Klensin's message of "Mon, 05 Apr 1999 12:44:51 -0400">  <199904040547.AAA05080@astro.cs.utk.edu>  <7EAG3p1mw-B@khms.westfalen.de>  <EXECMAIL.990405124451.P@SONY.mci.net> <4.2.0.32.19990405124055.00abc6e0@mail.imc.org>
From: Russ Allbery <rra@stanford.edu>
In-Reply-To: Paul Hoffman / IMC's message of "Mon, 05 Apr 1999 13:06:31 -0700"
Date: 05 Apr 1999 15:38:28 -0700
Message-ID: <yl3e2eofq3.fsf@windlord.stanford.edu>
Lines: 37
X-Mailer: Gnus v5.4.66/Emacs 19.34
List-Unsubscribe: <mailto:drums-request@cs.utk.edu?Subject=unsubscribe>

Paul Hoffman / IMC <phoffman@imc.org> writes:

> Read the rest of John's message. If an ISP allows a small amount of
> relaying, but less than what one would typically be called an open relay
> by spammers,

Er... an open relay is an open relay.  I agree with what people are
saying; there are good reasons for running an open relay under some
circumstances, there are ways to make them unattractive to spammers via
other means than closing them, and closing an open relay is a hard problem
in a lot of environments.  But it's still an open relay.  ORBS lists open
relays.  If people are uncomfortable with that level of aggressiveness,
they shouldn't use ORBS.

(I don't use ORBS.  I strongly dislike it, in fact, precisely for this
reason -- it lists people who don't spam.  But it does list only and
exclusively open relays, which is precisely what it says it lists.  It's
nothing if not honest.  If people want to reject mail on the basis of the
mechanical check outlined by ORBS, that's their lookout.  I have a policy
of not bothering to resend mail to people who block it for what I consider
to be stupid reasons.  If they want to communicate with me, they have to
avoid using dumb anti-spam techniques.)

> The problem is that ORBS is not actively maintained and therefore is
> assuming that no spammer is smarter than their mechanical setup.

Well, yes, there are ways of running an open relay that passes ORBS.  But
most of them are at least reasonably obscure.

> Also, there is the assumption that their mechanical setup catches only
> truly open relays.

Unless you have opened relaying to their specific probe host for some
reason, and I have to say that's unlikely, it pretty much does.

-- 
Russ Allbery (rra@stanford.edu)         <URL:http://www.eyrie.org/~eagle/>