That's definitely not the case with our systems: ----example---- X-Spam-Level: X-Spam-Status: No, score=-105.9 required=7.0 tests=ALL_TRUSTED,BAYES_00, HTML_MESSAGE,USER_IN_WHITELIST autolearn=ham version=3.0.1-prforsa3.0.0 ----end ex----
Something seems a bit fishy there. -----Original Message----- From: Jason Lixfeld [mailto:[EMAIL PROTECTED] Sent: Tuesday, November 09, 2004 1:44 AM To: users@spamassassin.apache.org Subject: SOLVED [Re: Reading, but not loading SQL userprefs (SA3.0.1 & Exim 4.43)] Figured it out. It seems that SA doesn't apply whitelist scores unless the rest of the scores add up to be greater than the score required to trigger X-Spam-Status: Yes. This seems to be a rather odd methodology. What's the rationale behind not taking into consideration whitelist scores regardless of the spam score? On 8-Nov-04, at 10:28 AM, Jason Lixfeld wrote: > I've got a whitelist_from in my SQL DB which Spamassassin is supposed > to read in. When I do a debug of spamd, I can see the connection to > the DB as the user, but the whitelist_from doesn't seem to be getting > applied to the message: > > mysql> select * from spamassassin.userpref; > +------------------+-------------------------+-------------------- > +--------+ > | username | preference | value | > prefid | > +------------------+-------------------------+-------------------- > +--------+ > | $GLOBAL | whitelist_from | [EMAIL PROTECTED] | > 36 | > | jason.lixfeld.ca | whitelist_from | [EMAIL PROTECTED] | > 35 | > | $GLOBAL | score USER_IN_WHITELIST | -10 | > 23 | > +------------------+-------------------------+-------------------- > +--------+ > 4 rows in set (0.00 sec) > > Below is the output of spamd in debug mode. Spamd is called twice for > this message. First to obtain the $spam_score and $spam_score_int > for data checking: