Jeff Chan jeffc@surbl.org wrote:
Thanks sm and Rob. I agree with sm that it would probably be better if end users did not contact us directly. Mail administrators and abuse desk folks should have a much better chance of understanding what's going on, so we would prefer to hear from them in the event of a false positive. As Rob mentions, FPs tend to be quite rare since we're trying to blacklist only hard-core spammers on SURBLs.
OK, that's what I will do. They will see our generic spam message and contact our helpdesk, and I will be able to tell from the system log what the problem was. The log will record the URI and that it matched a SURBL record. I'm glad I asked.
Regarding rejection, I realize I just popped up on this list without intro. Columbia University's mail system refuses about 1 million messages a day based on Spamhaus, NJABL, DNSBL, high Spamassassin score, and rules written in Mimedefang. SURBL looks like a good addition to our arsenal.
Logging in test yesterday and overnight suggests that SURBL will catch 50,000 messages a day that are not already caught before Spamassassin (which we run last). That's very worthwhile. Great!
Joseph Brennan Columbia University Information Technology