Reid Priedhorsky wrote:
> 7.0 is a pretty aggressive score for rejecting at SMTP time. I think that
> if you configure it to be like that, you are guaranteed to have some
> spurious blocking. Try relaxing the reject-at-SMTP score (I /dev/null
> mails at 22.0).
Uh... yeeaaahhh. Let's see, when I set those scores I never, in months,
saw a legit message over 6. I set my hard limit to 8. Then watched for
months again and never saw a message rejected. Meanwhile spam was slowly
edging downward and 8 was too high so it dropped to 7. Since then no blocked
mail. Even with SA's new crappy scoring you'll note it didn't get rejected at
SMTP, it was marked-and-passed.
22? Uh, 22!? Do you have any notion how much that is? From my logs over
the past 8 days or so here's the number of rejects based on scores 7 and over...
7-10: 84
11-20: 530
21-30: 617
31-40: 620
41+: 175
7 through 22 would be an additional 735 messages that I'd sift through in
my "unsure" folder. Sure, I'd still reject 1291 messages but 735 is a tad
higher than the, uh, about 30 I deal with in the same time frame, most of
which is spam. In fact until SA decided to inject crappy tests even the
"unsure" folder was nearly all spam.
Yeah, I can fix it. I get that. Point is that I shouldn't have to. It's
obvious those lists are broken so why were they even included much less at a
score so high as to trip even SA's on soft-limit?
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
