To:
"Hollenbeck, Scott" <shollenbeck@verisign.com>
cc:
"'Klaus Malorny'" <Klaus.Malorny@knipp.de>, ietf-provreg@cafax.se
From:
Jaap Akkerhuis <jaap@sidn.nl>
Date:
Mon, 09 Apr 2001 14:21:17 +0200
In-reply-to:
Your message of Mon, 09 Apr 2001 06:50:05 -0400. <DF737E620579D411A8E400D0B77E671D7508EF@regdom-ex01.prod.netsol.com>
Sender:
owner-ietf-provreg@cafax.se
Subject:
Re: Nameserver MUST HAVE IP
I disagree that the requirement should be removed completely, and reiterate my earlier suggestion that replacing the MUST with a SHOULD would be more appropriate. I personally opt for the SHOULD. jaap PS The whole discussion is a bit off-topic, or at best, deals with the policy of the registry how it creates the zone-file. We, the dutch registry also have the ``use ony glue when needed'' model. but in the (email based) ``rrp protocol'' with the registrar, they can give the IP numbers as well. We always check wether the nameserver actually works properly before we do anything. So whether the delegation is accepted, the nameserver is changed etc. etc., we first check. Thus far I have personally never seen the problem that nameservers became completely (as demonstrated by paf) without glue although it is rumored that it happened once in the last 15 years. But it is a problem inherent to the current DNS protocol and, as others have noted, it is not something which can be solved by a provreg protocol adequately. A clueless operator (even when thinking he is smart) can always manage himself in a glueless situation when he works hard at it. An, strictly speaking, the provreg protocol doesn't specify one should create a zonefile. That is outside the scope of this protocol. So even with a MUST in the requirement, it is still up to the policies of the TLD whether this information is actually used in the zone-file.