[copied off list message back to list]
Russ Ringer wrote:
On Fri, 09 Dec 2005 18:02:58 -0500, you wrote:
On 09/12/2005 5:52 PM, Justin Mason wrote:
Matt Kettler writes:
Really I think the use of notfirsthop in DUL testing is just plain broken. SA
should only be checking the host that drops off to your MX against the DULs. It
shouldn't be backtracking further.
To be honest, I'm inclined to agree -- I thought we *had* fixed that. :(
Is there a bug open about that?
http://issues.apache.org/SpamAssassin/show_bug.cgi?id=4728
Is this also affecting RCVD_IN_SORBS_PROXY as well?
This rule seems to have the same problem.
If by "RCVD_IN_SORBS_PROXY" you are referring to one of:
RCVD_IN_SORBS_BLOCK
RCVD_IN_SORBS_HTTP
RCVD_IN_SORBS_MISC
RCVD_IN_SORBS_SMTP
RCVD_IN_SORBS_SOCKS
RCVD_IN_SORBS_WEB
RCVD_IN_SORBS_ZOMBIE
then no. Each of the lists they check contain open proxies. It doesn't
matter where in the mail stream the open proxy is, it's still an open
proxy that will relay mail for anyone.
If you're seeing one of these hit in legit mail then slap the person
responsible in the head. In the event that it's a false listing, notify
SORBS. In either case, it's correct for SpamAssassin to check all
relays against these lists.
Daryl
