On Jun 16, 2006, at 4:41 PM, Hallam-Baker, Phillip wrote:
Defining a DKIM specific RR type should not prevent deployment.
Defining the DKIM specific RR within the base draft means a new
DKIM version is not required to enable its use. If there are
issues handling a DKIM specific RR, this is a problem independent
of DKIM. A new RR issue, if there are any, may delay exclusive
reliance upon the DKIM specific RR.
I expect the deployment of DNSSEC to solve the RR problem.
This seems to suggest DNSSEC clears the way for reliance upon EDNS
where RR size becomes less of an issue. Being pessimistic, the
relative timing for employing 2048 bit keys may not coincide with
developments in DNS. It would be good to get opinions on that topic
from the DNS group. Having a viable alternative not gated to
independent developments seems a more conservative approach. It may
be worth having a brief presentation of the process for publishing
and retrieving content of new RR types to get a better understanding
of current practices.
I believe DKIM creates the market for DNSSEC.
It will be interesting to see what is uncovered by either. Perhaps
DLV offers a solution for both. : )
-Doug
_______________________________________________
NOTE WELL: This list operates according to
http://mipassoc.org/dkim/ietf-list-rules.html