[dns-operations] was Re: on amplification, udp, and dns

[Edward Lewis]

At 8:20 +0100 3/24/06, Daniel Karrenberg wrote:

>The only path with real leverage is to prevent source address spoofing.

This is off-list because it's not related to DNS.

What makes me uneasy is that there's an interdependency (between 
routing and DNS) in the solution that shouldn't be there.

If we assume a protocol will be healthy only if the others around it 
are well-behaved, we have a "house of cards."  Things like SiteFinder 
and other changes (I'll not label them innovations, but the 
difference between an innovation like DNSSEC and SiteFinder is 
subjective) will suffer greater unintended consequences.  And 
piece-meal adoption of technology into other (non-Internet) 
environments will be problematic.

Perhaps it is a dream, but protocols ought to be able to live in "silos."

-- 
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Edward Lewis                                                +1-571-434-5468
NeuStar

Nothin' more exciting than going to the printer to watch the toner drain...