> At 12:42 AM 10/3/03 -0400, Christopher M. Iarocci wrote:
> >Hi, I'm using SpamAssassin 2.60 being called through Amavisd-new.
Dynablock
> >(along with a few others) is scoring on lots of messages that it
shouldn't
> >because the first hop is a dial up. I put the following line in my
> >/etc/mail/spamassassin/local.cf
>
> In 2.60 SA should ignore the first hop by default when doing dynablock.
>
> The standard RCVD_IN_DYNABLOCK rule already calls the eval with the
> "dynablock-notfirsthop" parameter.
>
> If it's not working, then it's a bug in the spamassassin code itself, and
> cannot be fixed by configuration options.
>
> I'd check the message very carefully.. are all the IPs beyond the first
not
> listed? Are you sure the 2nd, 3rd and 4th received lines contain IPs that
> aren't listed?
Absolutely. I checked, and re-checked all the IPs (there were 3) in the
header of the message I got, and none of them were listed in DYNABLOCK but
the originating one, yet it scored a 2.5 for DYNABLOCK.
>
>
>
> >header IGNORE_FIRST_IP_DYNABLOCK eval:check_rbl('foo-notfirsthop',
> >'dynablock.easynet.nl')
>
> Um... what is that supposed to do? You can't fix one rule by creating a
> different rule.
Just following the Mail::SpamAssassin::Conf page
>
> If it was an error in the rule, you'd have to change the RCVD_IN_DYNABLOCK
> rule to fix it.
>
> However, this is the rule, and it's already "fixed" so the bug lies in the
> code itself, if it exists.
I'd have to say that a bug exists then. Also, the bug seems to exist with
multiple RBLs, not just DYNABLOCK as the message hit NJABL and NJABL_DIALUP
which I also checked, and again, none of the IPs but the originating one was
listed.
Christopher M. Iarocci
-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Spamassassin-talk mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
