--On June 19, 2006 6:06:54 PM -0400 Micah Anderson <micah@riseup.net> wrote:
Since we are talking bad blacklisting... we just got blacklisted by Spamhaus because they felt as if the people providing us our upstream connection weren't dealing with the spammers on their network to their satisfaction. This meant a overly-broad block which included our class C (they blocked a /20), even though we demonstrated to them that we do not have a problem and actually have spam policies that we act aggressively to enforce. Fortunately, this has been resolved, but I am surprised that Spamhaus doesn't mind taking on some collateral damage.
BZZZZZZZT! Spamhaus, et al, block *NOTHING*. They publish lists of information (not necessarily facts), and, in some cases yes, opinion. The people doing the blocking are those electing to use this information. And some use it very poorly. We keep automated denial usage to a minimum, and only use lists with clearly defined addition and removal policies that have no humans in the loop (except in whitelisting cases). We don't use things like spamhaus and sorbs at the MTA level because of the problems you mentioned. We do allow points to be applied in later filtering via SpamAssassin though, but a hit on spamhaus alone isn't nearly enough to trigger a message as spammy unless you intentionally turn point threshold so low as to get a lot of falses.
Thus your problem isn't so much with the list maintainers, as it is with those using the lists in a draconian manner.
-- Michael Loftis Modwest Operations Manager Powerful, Affordable Web Hosting