Hi,
I'm experiencing the exact same problem with 3.2.3, y "fix" was simply
to manually specify the spam score:
# adjust for high efficiency rules
score URIBL_BLACK 50
score URIBL_JP_SURBL 50
score RCVD_IN_BL_SPAMCOP_NET 50
score RAZOR2_CHECK 50
score BAYES_99 50
# short circuit high efficiency rules
shortcircuit URIBL_BLACK spam
shortcircuit URIBL_JP_SURBL spam
shortcircuit RCVD_IN_BL_SPAMCOP_NET spam
shortcircuit RAZOR2_CHECK spam
shortcircuit BAYES_99 spam
-- Felix
Crocomoth schrieb:
Hello,
I am using version 3.2.5 and, to reduce spam and SA footprint, turned on
shortcircuit plug-in.
I use standard 60_shortcircuit.cf.
At the first time, I enabled only whitelist/blacklist rules and this works
great.
Now, due to increasing amount of spam, I decided to turn on bayes
shortcircuit rules.
Shortcircuit doc
(http://spamassassin.apache.org/full/3.2.x/doc/Mail_SpamAssassin_Plugin_Shortcircuit.html)
says that in case of "spam" it "override the default score of this rule with
the score from shortcircuit_spam_score" but this was not true - plug-in does
not change score and it remains 3.5 which is lower than my threshold and it
breaks further checking and the situation gets worse than without this rule
- I have started receiving much more spam (with score 3.5 from bayes).
Ok, I have added to this file "shortcircuit_spam_score 100" and
"shortcircuit_ham_score 100" explicitly, but this did not help - result was
the same - looks like these options do not work at all.
Ok, I think, it is possible to change bayes_99 score in appropriate file,
but, I think, this will not be correct.
Thanks.
