[mail-vet-discuss] New draft for review

[J.D. Falk]

On 2007-05-29 14:25, John Levine wrote:

>>> I don't see the advantage of having it say
>>> in-23.atl.mail.earthlink.net rather than mail.earthlink.net.
> 
>>Agents use the identifier, the FQDN in this case, to determine 
>>whether the Authentication-Results header can be trusted.  The FQDN 
>>may not the best choice in the case of mail farms.
> 
> Well, yeah.  Seems to me that mail farms are likely to apply the
> majority of these headers, so we better have something that works
> with them.

Stepping back for a moment...is there any reason that the identifier 
can't be 'icannhascheezburger' or some other nonsense, so long as the 
internal network (and perhaps MUAs which connect to that internal 
network) knows what to look for?

If not, then perhaps the answer is that the identifier SHOULD be unique 
to that network, and it is RECOMMENDED to use the hostname (either real 
or virtual) of the MTA which determined the contents of the header.

And, of course, MTAs SHOULD ignore that identifier on messages coming 
from outside of their trusted network -- but as John pointed out, the 
trusted network may easily be large and amorphous.

-- 
J.D. Falk, Anti-Spam Product Manager
Yahoo! Mail