Re: r= for instilling good domain-name practices

On Sat, 2006-04-29 at 14:27 +0000, John Levine wrote:

> >This seems to me to be a poor man's version of SSP, ...
> 
> The whole issue of mail categories is an interesting one, but it's a
> large, deep, chewy can of worms relevant to much more than DKIM that
> we'd be nuts to try to address here.
> 
> Those of us who are nuts are welcome to reconvene in the ASRG where
> it's already a topic of occasional discussion.

Once the r= parameter is established as an option in the DKIM mechanism,
conventions for what is implied by the different levels of r= parameter
can, and perhaps should, be defined by other groups.

It is however vital that some standardize convention for expressing this
information be provided by the DKIM mechanism, or there is a real
possibility that DKIM will make the situation worse when signing domains
resort to using more domain names to distinguish between differ levels
of vetted sources.  This strategy would greatly weaken domain-name
recognition, and much of the value that could be derived by
incorporating DKIM signing. 

-Doug