Hello Rob,

Sunday, August 15, 2004, 10:41:08 PM, you wrote:

RB> I have been running SA for about 6 months without out much trouble,
RB> having it catch 98 or 99% of my spam (20-40 a day), but recently, I
RB> have run into some problems with the Bayesian filters getting
RB> corrupted, and digging around for some new rules, I adjusted my
RB> rules, and now all sorts of easy to catch spam are not getting
RB> caught.

RB> I am invoking SA (spamassassin-3.0.0-0.pre4.2.3mdk)  from Kmail, by
RB> passing it through spamc running rules_du_jour and my_rules_du_jour
RB> and using the rules: SARE_RATWARE, SARE_BAYES_POISON_NXM, SARE_SPOOF,
RB> SARE_OEM, BOGUSVIRUS, SARE_ADULT. Now, not only is this list getting
RB> a bit unweildly, taking about a minute to process all uploads, but it
RB> is doing a crummy job of filtering.

I'd suggest adding SARE_GENLSUBJ-0 and 1, SARE_HEADER-0 and 1, and
SARE_HTML-0 and 1.

Are your network tests working?  Are the various network blacklists,
especially the new SURBL, doing their job? Losing the network checks
could have this type of effect.

Bob Menschel



Reply via email to