> Should I see this for mail that is scanned for spam? If so, am I > missing something in configuring this to be used? watch the output of "spamassassin --lint -D" ... debug: bayes corpus size: nspam = xxxx, nham = xxxx Uwe