If I have to send one more of these "oops" messages I may just set up a
template for them.

I've been complaining about false positives due to high Bayes values. My
training script had an error in it and my largest source of ham was being
trained as spam.

I think tihs is classified as a PEBCAK bug. (Problem Exists Between Chair
And Keyboard) My apologies for the confusion I've caused.

I still think the BAYES_99 should be below the cutoff, but I may change my
mind after watching this run with untainted Bayes data.

-- 

Jack Gostl      [EMAIL PROTECTED]



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Spamassassin-talk mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/spamassassin-talk

Reply via email to