We have recently noticed that now we are getting more and more Negative scores from Spam assassin on the Ubuntu804. 1. Is this normal? 2. A good amount of spam is showing up with these negative numbers and is making us a bit less than efficient TP
Seems to be in the Bayes When I look through the messages that were getting the negative scores, Bayes_00 was setting the scores to like -4.0 and -3.0 So when we shut Bayes off, scores come back into line...however sa learn is toast. ideas? TP
Hey, Can you tell me what you have as your Spam Lists setting in MailScanner.conf? Also, when you say that sa learn is toast, do you mean that it's not working? You can updated spamassassin by logging into your MailWatch interface and going to Tools. Theres an option to update spamassassin.
Sorry...been delayed a bit Back to the issue. I had to shut Bayes off to get this to work and actually filter spam...other wise -7 scores were the norm..and spam was getting through left and right. When I shut down Bayes, we got much much better. So obviously we would like to use Bayes....and hence get the SA Learn capability back. To Answer you direct question in /etc/MailScanner/MailScanner.conf we have: Spam List = spamcop.net SBL+XBL All help is appreciated!! TP