Re: Evaluation: draft-ietf-dnsext-unknown-rrs - Handling of Unknown DNS Resource Record Types to Proposed Standard

[Rob Austein <sra@hactrn.net>]

At Sun, 11 May 2003 22:28:48 -0400, Steven M. Bellovin wrote:
> 
> It's clearly necessary to have something like that, but frankly, the 
> document scares me; it retroactively changes the behavior required for 
> older RFCs.  I sure with Mockapetris had thought of saying this.
> 
> Am I offbase?  Is this much better -- or much worse -- than I fear?

The retroactive changes are mostly DNSSEC-related, and the DNSSECbis
drafts take the same approach.

The scariest non-DNSSEC part of this draft is presumably the name
compression stuff in section 4, but note (as this draft does), that
this section is basicly just finishing up a job that RFC 1123 started
(section 6.1.3.5, to be precise).

So I agree that it's scary, but I don't know how to do significantly
better without a time machine.