Re: Mailing lists - assumptions

On 4/19/2014 11:45 PM, Scott Kitterman wrote:
>>
>> People might read the about as messages which pass DKIM or SPF will
>> not be rejected.  That's not what those mail receivers are doing.
>
> Right. I think it bears repeating for emphasis that the current Yahoo! DMARC mess is unrelated to issues with SPF as defined in either RFC 4408 or 4408bis.
>
> Typical MLM managed lists send messages that pass SPF for the list's domain (which is typically what is found in Mail From).  Conceptually this is somewhat similar to the list adding it's own DKIM signature. Both pass/verify for SPF/DKIM, but neither are aligned to the From, so DMARC calls them a fail.
>
> SPF and transparent forwarding is an issue, but not this one.
>

Right. Scott, can you confirm my understanding of the DMARC policy 
yahoo.com has and how does it describes any tight alignment?

     "v=DMARC1;
        p=reject;
        sp=none;
        pct=100;
        rua=mailto:[email protected],
            mailto:[email protected];"

According to this, there is no strict alignment.  The default 
alignment is relaxed adkim=r aspf=r.

So doesn't this say?

     "All mail from YAHOO.COM must be signed, but we don't care who.
      If its not signed, reject it."

Why would transactions from a yahoo.com subscriber to yahoo.com 
subscribers get rejected, for example, with these data bits:

     5321.FROM FROM:      [email protected]
     5321.RCPT TO:        [email protected]
     5322.From:           [email protected]
     5322.DKIM-Signature: d=winserver.com

The reject response is:

     5.7.9 Message not accepted for policy reasons.
           See http://postmaster.yahoo.com/errors/postmaster-28.html

Where is the alignment requirement description for when only p=reject 
is used?

-- 
HLS


_______________________________________________
ietf-822 mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/ietf-822