http://bugzilla.spamassassin.org/show_bug.cgi?id=2990





------- Additional Comments From [EMAIL PROTECTED]  2004-01-31 13:10 -------
Subject: Re:  proposed interim rules for 2.64

bugzilla-daemon  <[EMAIL PROTECTED]> writes:

> -1 on this.  I'll be perfectly happy with:
>
> score HABEAS_SWE -1 -8 -1 -8
>
> but not -1 across the board -- if VIOLATOR can be hit (ie: net-enabled
> score set), we should leave the score as-is.

VIOLATOR can be hit, but it isn't being hit.  I have 15 hits for
HABEAS_SWE in my missed spam folder and 0 HABEAS_VIOLATOR hits.
HABEAS_VIOLATOR is completely ineffective.  It's responsible for 10% of
my missed spam with 2.6x.

For caught spam, I have 27 HABEAS_SWE hits and only 14 HABEAS_VIOLATOR
hits.  That also is not very good.

When I changed the score from -4 to -8, it was on the premise that the
rule was working for SpamAssassin users.  In the face of compromised
open proxies, I think the honeymoon is over, this rule is no better than
any simple header negative rule we've ever had.

A score of 0 is probably better than -1 for non-network score sets and
is consistent with our policy against easily-forged rules.

Daniel





------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

Reply via email to