On Sat 19/Mar/2016 00:00:18 +0100 Franck Martin wrote: 
> ----- Original Message -----
> 
>> From: "Tomki Camp" <[email protected]>
>> To: "Franck Martin" <[email protected]>
>> Cc: [email protected], "Alessandro Vesely" <[email protected]>
>> Sent: Friday, March 18, 2016 2:01:26 PM
>> Subject: Re: [dmarc-ietf] SPFAuthResultType unbounded
> 
>> Ok, I can do that.
>> I'd like a further suggestion on how to specify more detail for this change:
> 
>> - <xs:element name="spf" type="SPFAuthResultType" minOccurs="1"
>> - maxOccurs="unbounded"/>
>> + <xs:element name="spf" type="SPFAuthResultType" minOccurs="1"
>> + maxOccurs="2"/>

Why not maxOccurs="1", after Franck's note below?  Since most feedback
producers supply one SPF result, I'd rather seek homogeneity than introduce
further intricacies.

>> Because if there are more than just 1, both of the possible SPFAuthResultType
>> entries need to indicate the scope, and they must be different. (i.e. one
>> must be mfrom, the other must be helo)
> 
> If not mistaken, in the spec, we say DMARC uses only the mform evaluation
> (rfc5321.mailfrom or rfc5321.helo if the first one is empty).

Exactly, see https://tools.ietf.org/html/rfc7489#section-4.1

The possibility to use helo if mfrom is void is described in the second
paragraph of http://tools.ietf.org/html/rfc7208#section-2.4

Ale
-- 

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

Reply via email to