Bah I need to quit typing so much. (you'de never guess I've been on the list since the beginning of the year)
I agree with you and realize how effective it is. ( started warn_if_reject yesterday afternoon on a minimal subscriber regexp. 5k warnings last night 3k+ this morning by 9am ) The 1% of legit people on a DSL/cable need to get off their high horse and relay through their ISP. Its not worth all the spam to not use such a great filter. Len: do you share your subscriber_networks.regexp as part of basic or advanced? Hate to spend all this time building mine if you share yours openly. -----Original Message----- From: Len Conrad [mailto:[EMAIL PROTECTED] Sent: Monday, September 15, 2003 12:09 PM To: [EMAIL PROTECTED] Subject: [IMGate] Re: Charter - subscriber blocks >(charter) from=<[EMAIL PROTECTED]> >to=<[EMAIL PROTECTED]> cpe-24-107-232-14.ma.charter.com Now you see why I expected that Scott would get on my case so hard about this recommendation in the Imail list, and he did. His attack on my recommendation smells like a self-serving hidden agenda. I doubt there's any good reason he can't relay declude.com's outbound through his won ISP of Computerized Horizons, since that would go along with my standard recommendation that legit mailers on subscriber networks should do just that. >I've since excluded them... But is there a known good exception for >(other) businesses with a connection via charter? That specific one is probably the only one common to this group. But don't buy my "preaching" as Scott calls it, just run in warn_if_reject for a few days, identify any subscribers IPs that habitually send to your MX (probably on the order of 1 in 1000 subsciber IPs that abuse you), whitelist them (or prabably, "it"), and then move to reject mode. You will still be way ahead with this cautious approach vs not doing it at all. And I assume that you have tested all your relay domains for all incoming mail from Internet to postmaster@ and abuse@ actually showing up in somebodys' mailboxes. This practice is of course not specific to the subscriber filter. Len
