Re: Last Call: Root Name Server Operational Requirements to BCP

[Scott Bradner <sob@harvard.edu>]

> As has been pointed out in the past, this  document is likely mis-named.

I don't understand the comment - it sounds like Bill is saying
that the doc is misnamed because it only covers what the name says
it covers

Scott

----

From scoya@ietf.org Sun Mar 12 03:21:12 2000
From: Bill Manning <bmanning@ISI.EDU>
Subject: Re: Last Call: Root Name Server Operational Requirements to BCP
To: randy@psg.com (Randy Bush)
Date: Sun, 12 Mar 2000 00:20:52 -0800 (PST)
Cc: djb@cr.yp.to (D. J. Bernstein), iesg@ietf.org, dnsop@cafax.se
In-Reply-To: <E12U2ij-000Dju-00@rip.psg.com> from "Randy Bush" at Mar 11, 2000 11:23:25 PM
X-Mailer: ELM [version 2.5 PL2]
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Transfer-Encoding: 7bit

% 
% > Perhaps you'll say that you don't care whether all the TLDs are broken;
% > you're just worrying about the root zone.
% 
% correct
% 

As has been pointed out in the past, this  document is likely mis-named.
And I believe that simply focusing on root servers is wrong. The "trickledown"
effect is pervasive in the DNS architecture.  RFC 2010 talks about rootserver
expectations and recognises that at the time, there was a co-mingling of
root and TLD service on the same suite of servers. It also comprehended that
the TLD service should be moved to their own sets of servers. This document
seems to retain the view that root and gTLD zones will remain comingled. I
think it is better addressed to the "flat" gTLDs with large numbers of zones.
In any case, I think that Randys focus on just the root is misguided. It is 
important to have compatability betwen parent and child servers. And this 
documents label does not reflect that.

-- 
--bill