Re: [Asrg] Email service assumptions and making system-wide changes

[Dave Crocker <dhc at dcrocker.net>]

> Yeah, mail should be permited/forbiden based upon the relationships between
> sender and recipient, but might be actioned at any point.

I think this is quite a good and powerful summary statement.  One worth re-using 
quite a lot, though not as the sole, absolute directive for all anti-abuse/spam 
efforts.

I believe there are two, complementary sets of relationships worth focus, the 
one you cite and the one between operators.

One of the things that is interesting about DKIM is that it is mostly discussed 
as being between operators but it really is between originators and recipients, 
albeit without the final granularity of mailboxes -- it uses only the coarser 
domain name reference.  That it can be *administered* by operators is a very 
useful choice, but the fact that it does not *require* operator participation is 
fundamental.  Contrast that, for example, with SPF. Another contrast would be 
CSV, which is strictly between neighboring operators.

d/

ps. I am making the assumption that your use of "sender" means either 
rfc2822.From or rfc2822.Sender.  In general, the raw use of the word "sender" 
has become highly ambiguous, given how many intervening agents are senders.

--

Dave Crocker
Brandenburg InternetWorking
<http://bbiw.net>

_______________________________________________
Asrg mailing list
Asrg at ietf.org
https://www1.ietf.org/mailman/listinfo/asrg