Hi, I think I messed up my bayes db. When I first implemented Spamassassin 2.63, I didn't train it correctly. I started training it with a lot of spams and a few hams. Here's my sa-learn --dump magic result.
0.000 0 2 0 non-token data:bayes db version 0.000 0 9657 0 non-token data:nspam 0.000 0 203 0 non-token data:nham 0.000 0 130984 0 non-token data:ntokens 0.000 0 1089255579 0 non-token data:oldest atime 0.000 0 1092157872 0 non-token data:newest atime 0.000 0 1092158029 0 0 non-token data:last journal sync atime 0.000 0 2764800 0 non-token data:last expire atime delta 0.000 0 34190 0 non-token data:lastexpire reduction count I start to realize there's something wrong when I don't see those data change, even when I have set auto_learn 1 in local.cf. Then I try the sa-learn --rebulid and sa-learn --force-expire. Those data don't change neither. Is there a way to forget all the 9657 nspam and re-train the spam again? or I have to reinstall spamassassin from start? any suggestions? Thanks. -- Eleanor Cheung __________________________________ Do you Yahoo!? Yahoo! Mail - 50x more storage than other providers! http://promotions.yahoo.com/new_mail
