https://issues.apache.org/SpamAssassin/show_bug.cgi?id=5684

[email protected] changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |software+spamassas...@kd6lv
                   |                            |w.ampr.org

--- Comment #15 from [email protected] 2010-11-15 16:27:59 
UTC ---
I am of the opinion that this is a poor approach.  The condition should not be
whether the SPF record contains a "+all" but that such a mechanism is the
reason why a "pass" result is generated.

Consider:  "v=spf1 a mx all".

Clearly, if the "a" or "mx" mechanism is the reason for the passing result,
there is nothing "funny" going on here, despite having a "+all" in the record.

Also, SPF is better done at the SMTP "MAIL FROM" state, not the "DATA"/eom
state where the message has already been transferred.  SPF failures should be
rejected earlier in SMTP.  However, SPF "none"* vs. "neutral" vs. "softfail"
are still results for SA scoring.  (* - see bug 6490 for "none" being available
to SA.)

RE - Comment #7:  I agree, and I also disallow by policy any passing result
that came from a CIDR'ed mechanism where the mask is less than 8.  I chose "/8"
because: 1)  It is the largest allocation to any single organization, and 2) it
divides the IPv4 address space into 256 blocks - and for a spammer to cover
enough of it to be worthwhile, he would violate SPF's mechanism limit and/or
the DNS size limit on SPF or TXT RRs, even with nested included SPF records.

I call SA from MIMEDefang, so I perform my primary SPF check in MD's hook
filter_sender() routine.  When it calls Mail::SPF, one of the returned fields
is an explanation of the result.  When pass results, the explanation is usually
"Mechanism '*' matches", and I parse the "*" for a CIDR or "all". Such "wide"
CIDR-matching pass results are changed to the "policy" result per RFC 5451.

Such action makes this patch completely unnecessary, plus it targets the wrong
criterion.

-- 
Configure bugmail: 
https://issues.apache.org/SpamAssassin/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

Reply via email to