Quote:
required_score 4.5
rewrite_header 1
subject_tag 1
whitelist_from xxxx xs4all.nl
whitelist_from xxxxxs4all.nl
blacklist_from *.bravenet.com
|
Code:
rewrite_header subject **MOGELIJK SPAM**
# rewrite_header is set by entering a value only,
# it does not accept a true/false 1/0
# you need to specify which header to rewrite,
# ie, subject (you can rewrite to, from, or subject header)
delete "subject_tag 1" not used in SA v3
blacklist_from *.bravenet.com
# --- *. forget the . use just *
# (*=everything) --- *bravenet* also gets .net, .org, .etc
# But, only by adding -100 to the score
# When using cPanel filtering for SpamAssassin, it has to be the last filter
# in the list or subsequent filters could also process the e-mail.
While this is working, I think you are making SPAM detection a bit complicated and time consuming. I just do not have the time to parse scores and deal with bayes learning.
I will share with you my user_pref file which currently is yielding zero spam. I
DO use a filter file later to allow my whitelist deliveries, redirection of matching criteria, and then blackholes all marked SPAM and non-spam blacklists. Spamassassin is not the fool-proof solution to white and black lists, that is done in the filter file.
Code:
required_score 4
rewrite_header subject **_HITS_**
use_bayes 0
score HTML_FONT_SIZE_TINY 4.0
score BEST_PORN 4.0
score HTML_IMAGE_ONLY_04 4
score HTML_IMAGE_ONLY_08 4
score HTML_IMAGE_ONLY_12 4
score HTML_IMAGE_ONLY_16 4
score HTML_IMAGE_ONLY_20 4
score HTML_IMAGE_ONLY_24 4
score HTML_IMAGE_ONLY_28 4
score HTML_IMAGE_ONLY_32 4
score HTML_TINY_FONT 4.0
score MALE_ENHANCE 4.0
score PORN_16 4.0
score URIBL_JP_SURBL 4.0
score HTML_FONT_TINY 4.0
score STOCK_ALERT 4.0
score MIME_BASE64_TEXT 4.0
score US_DOLLARS_3 4.0
score URIBL_SBL 4
score RCVD_IN_BL_SPAMCOP_NET 4
score STRONG_BUY 4
score DRUGS_ERECTILE 4
score FUZZY_REFINANCE 4
score HTML_10_20 4
score HTML_MIME_NO_HTML_TAG 4
score SUSPICIOUS_RECIPS 4
score RCVD_IN_SORBS_DUL 4
score RCVD_IN_NJABL_PROXY 4
score URIBL_OB_SURBL 4
score UPPERCASE_25_50 3.5
score UNPARSEABLE_RELAY 4
score HTML_FONT_FACE_BAD 3.5
Viewing your message headers, SpamAssassin documents which tests it used to calculate the score. To see the list of tests and their default scores, go to
http://spamassassin.apache.org/tests.html . I set the score of the definite rules I don't want to my required SPAM score.