Hi folks,
I have a strange problem on my NetBSD-current box:
I installed spamassassin from pkgsrc (V 3.2.3) and configured it in
sendmail (INPUT_MAIL_FILTER(`spamassassin'...).
It seems to work; there are bayes_seen and bayes_toks, and I can see
in `sa-learn --dump magic` my spamassassin seems to learn.
But newly arrived mail gets never marked as spam; spamassassin[1]
checks it, but the header rewrite is unusual:
Mar 25 18:05:06 vinyl spamd[4661]: spamd: processing message
<[EMAIL PROTECTED]> for me:1005
Mar 25 18:05:06 vinyl spamd[4661]: spamd: clean message (0.0/5.0) for
me:1005 in 0.1 seconds, 4399 bytes.
Mar 25 18:05:06 vinyl spamd[4661]: spamd: result: . 0 -
scantime=0.1,size=4399,user=me,uid=1005,required_score=5.0,rhost=localhost,raddr=127.0.0.1,rport=64556,mid=<[EMAIL PROTECTED]>,autolearn=ham
Mar 25 18:05:06 vinyl sm-mta[914]: m2PH54SR000914: Milter add: header:
X-Spam-Checker-Version: SpamAssassin 3.2.3 (2007-08-08) on myhost
Huh? The only header is "X-Spam-Checker-Version", but it does not
write a "X-Spam-Status" or "X-Spam-Level" header to my messages as I
can see in message source. Isn't that strange? Exactly the same setup
worked fine under a Debian Xen domU...
`spamassassin -D --lint` without any errors.
I hope anybody can give me a hint.
Best regards,
Marianne
[1] required_score 5.0
rewrite_header Subject [:.SPAM.:]
report_safe 1
use_bayes 1
bayes_path /var/spamassassin/bayes
bayes_file_mode 777
bayes_auto_learn 1
bayes_auto_learn_threshold_nonspam 1.0
bayes_auto_learn_threshold_spam 8.0
--
"Die einen nennen es 'Zensur', die anderen 'regular expression'."