Fishtop_records wrote:
I've got the rest of the rules based stuff working fine, but I can't seem to
get the bayesian stuff scoring.
When I run sa-learn --debug magic, you can see that there are lots of graded
messages
0.000 0 1224 0 non-token data: nspam
0.000 0 1280 0 non-token data: nham
What user did you run that as? Are you sure its the same user that's
invoking spamc? (and also not root.. spamd never scans mail as root)
and there are no errors in the syslog.
spamd and spampd report no errors to syslog (it took a while to get the
complaints about writing the bayes lock to be fixed)
