On 12/9/07, Cyndi Norwitz wrote: > But this isn't useful to me. Oh, I'm sure some of the really bad spam > would go away, but this is a health list and so there are *a lot* of false > positives because we mention a lot of spam-like keywords. So I'd have to > set the spam level pretty high.
It depends on how your ISP runs SpamAssassin. It is possible to run it in a manner where the user has full control over what rules and what scores will be applied to their mail, and when you train SpamAssassin by feeding it examples of spam that has gotten through or ham that has accidentally been mis-identified, these rules and scores will be updated as necessary. I've been seriously fighting spam for about twelve years (as the Sr. Internet Mail Administrator for AOL, I wrote some of the earliest comprehensive anti-spam measures for sendmail, which I then re-published to the community), and I've spoken on this subject at conferences, I've been a member of the IETF/IRTF Anti-Spam Research Group, and was the head of the Best Current Practices sub-group. I can tell you, with some authority, that the only effective way to run SpamAssassin is to do so using these per-user methods. And that if you (the ISP) do actually run it in this way, you really can quite effectively catch or identify most spam, even in environments where you would otherwise tend to generate excessive false positive matches. Of course, that doesn't mean that your ISP is actually going to do any of these things. > Here's what I want: > > Subscribers who are unmoderated to be whitelisted. > Non-subscribers who I have set to auto-accept to be whitelisted. That doesn't work, either. Spammers troll the archives of mailing lists to find addresses they can use -- to spam those mailing lists, among others. You can't just auto-whitelist all addresses in certain classes. > Potential spam from the moderated box to be sent to my graymail (my ISP's > name (or maybe a common name, I don't know) for suspected spam--they send > an email each night with the from and subject headers). No, that's not a common name. I've been in this business for nearly twenty years, and in all that time, I have never heard this particular term used in this manner. More common terms are folders with names like "quarantine" or "probable spam". -- Brad Knowles <[EMAIL PROTECTED]> LinkedIn Profile: <http://tinyurl.com/y8kpxu> ------------------------------------------------------ Mailman-Users mailing list Mailman-Users@python.org http://mail.python.org/mailman/listinfo/mailman-users Mailman FAQ: http://www.python.org/cgi-bin/faqw-mm.py Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/ Unsubscribe: http://mail.python.org/mailman/options/mailman-users/archive%40jab.org Security Policy: http://www.python.org/cgi-bin/faqw-mm.py?req=show&file=faq01.027.htp