http://bugzilla.spamassassin.org/show_bug.cgi?id=3295





------- Additional Comments From [EMAIL PROTECTED]  2004-04-22 17:32 -------
I've decided that the orbitz.com example is working as designed. Since
orbitz.com is not listed as trusted, there is no way to know that the Received
header that mentions mx.orbitz.com is not forged. If mail from [EMAIL 
PROTECTED] goes
first through mailfoo.wm.orbitz.com and then through mailbar.orbitz.com, the
correct domain to list in the def_whitelist_rcvd has to be orbitz.com. I could
see including both just in case.

Regarding the two remaining FPs: I'll deal with the second one first, because it
raises fewer questions.

If I interpret your obfuscation of names correctly, has the recipient at
example.net on a system that runs its mail server in the domain example.org.
example.org should be on the list of trusted_networks in this configuration, and
then I think the FP should go away. If it isn't listed as trusted, this system
is going to get other false positives for dialup rules, for example. Can you
confirm if this is correct?

The remaining FP is from a mailing list. It comes from a yahoo-inc address with
the last trusted server being that of the mailing list server. Now I don't have
a simple answer. How are we dealing with cases like this in other situations
where we have to accurately identify which Received headers can be trusted?




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

Reply via email to