On Tue, Aug 07, 2001 at 03:33:36PM +0100,
Alan J. Flavell <flavell@???> is thought to have said:
> criterion of actually having participated in the relaying of spam, as
> the MAPS RSS insists on. Hence the increased likelihood of what in
> practical terms we would call false positives (i.e the rejection of
> good mail on suspicion of being offered by a spam relay) - sure, in
> objective technical terms these aren't "false" positives at all: the
> system is working as designed (blocking mail, irrespective of its
> nature, from sites that are objectively proven to be open relays).
True. None of the services which popped up in the wake of ORBS operates as
RSS does. But then AFAIK they all are fully automated while RSS has a
level of manual intervention to not test servers unless it really is spam
that was submitted with the nomination. Having switched from using RSS to
block to ORDB to block I haven't noticed an appreciable change in the
number of "false" positives. But I am catching far more relay spam than
with RSS and having many fewer to submit to ORDB than I used to submit to
RSS.
> > So there is no chain there.
>
> I said "the address is (part of) an open relay (chain)".
>
> Are you saying that the address would be blacklisted for even
> accepting the mail, irrespective of whether it relayed it (whether
> directly or via a chain)? That would seem excessive, and doesn't
> appear to be what the site's own description says.
No. I misinterpreted what you said to be that all of the replacements
for ORBS listed the endpoints of the chain of servers involved in a
multi-hop open relay. Sorry about that.
Tabor
--
--------------------------------------------------------------------
Tabor J. Wells twells@???
Fsck It! Just another victim of the ambient morality
