>You can disable this test by adding the following line into your >local.cf or ~.spamassassin/user_prefs file: >score HABEAS_SWE 0 >or take a more moderate road and simply reduce the score: >score HABEAS_SWE -1.0
I have to admit I have NOT followed the huge thread on this, so if this suggestion was already made I apologize. The modest step I took was to create a local meta rule where if HABEAS_SWE and BAYES_99 are both true, then add 8 points to negate the -8 score from HABEAS_SWE. In effect, that disregards the HABEAS_SWE check IF bayes scores 99. My feeling (untested) is if this meta rule results in a few false positives, a few feedings into bayes of the FPs will drop the bayes score down a bit, since the headers are part of the bayes score, and legitimate mailers do not typically vary their sending address and mail servers on a constant basis. #-- From the local.cf -- #HABEAS_SWE is ham indicator with a score of -8. But it can be forged. #This is a meta rule so that if a message has both HABEAS_SWE and #BAYES_99, we assume Bayes is correct and add a score of +8 back to the #message to negate the HABEAS_SWE rule. meta HABEAS_SWE_BAYES_99 ( HABEAS_SWE && BAYES_99 ) score HABEAS_SWE_BAYES_99 8 -------------------------- John C. Ring, Jr. [EMAIL PROTECTED] Network Engineer Union Switch & Signal Inc.
