Hi Jared,
Do you mind if I redirect the below back onto the spamassassin list
and respond to it there?
I'm concerned that I might have a configuration error if a DNSBL
check was done against an IP from a Received header that wasn't the
last external one, as you mention.
Thanks,
Andy
On Thu, Sep 28, 2023 at 12:02:47AM -0400, Jared Hall wrote:
On 9/27/2023 5:42 PM, Andy Smith wrote:
> Hi Jared,
> > > Just under that paragraph I mentioned that the supplier hosts public
> mailing lists where their employees end others use addresses that I
> can't predict, so I think this would not work as I'd have to
> allowlist all mail from all of mimecast's clients.
> > Thanks,
> Andy
Not sure what all that means but it sounds like a SPF/DKIM nightmare.
Crazy
suppliers, but what can you do?
SpamAssassin doesn't arbitrarily pick a header to look at. lastexternal is
used per the defaults in 20_dnsbl_tests.cf
Header line ordering is a function of the milter. You didn't say what
milter you are using. You'll need a Spamass-milter patch to correct the
ordering of headers. Amavisd-new, and MimeDefang work fine as is.
The old
commercial milter-spamc preserved the header order correctly, but puts the
spamassassin results at the end (bottom) of the existing headers.
That's all I have for now.
I wish you good luck,
-- Jared Hall
