Re: [alac] Redirection of non-existing domain names

[Thomas Roessler <roessler@xxxxxxxxxxxxxxxxxx>]

While I'd certainly agree that faking DNS records about non-existing
domain names is a bad practice for registries to apply (and should
be banned), I don't think we should get into technicalities of that
topic in any advice we give -- that's what ICANN has the Security
and Stability Advisory Committee and the Technical Liaison Group (or
even directly the IESG/IAB) for.  They can make these statements
with much more weight than we can, even if we make them as correctly
and precisely as you suggest.

I think we should phrase our observations on this rather as a
concern which should be dealt with (because it affects Internet
stability and interoperability, breaks the expectations of users,
etc.), than as "final" advice.
-- 
Thomas Roessler                         <roessler (at) does-not-exist.org>






On 2003-05-28 12:08:18 +0200, Vittorio Bertola wrote:
> From: Vittorio Bertola <vb@xxxxxxxxxxxxxx>
> To: Wendy Seltzer <wendy@xxxxxxxxxxx>
> Cc: Interim ALAC <alac@xxxxxxxxx>
> Date: Wed, 28 May 2003 12:08:18 +0200
> Subject: Re: [alac] Redirection of non-existing domain names
> X-Spam-Level: 
> 
> On Wed, 21 May 2003 12:27:26 -0700, you wrote:
> 
> >Seconded.  
> 
> I have finally had some time to draft a text - comments welcome. I
> would then submit it to the Board (via the CEO?), to the GNSO Council
> and to the ccTLDs, and copy it to the GAC (perhaps also the SAC?).
> 
> ====
> ALAC Statement on Redirection of Unregistered Domain Names
> 
> The ALAC would like to express its worry for the growing practice of
> providing affirmative replies to DNS queries for unregistered domain
> names at the registry level.
> 
> We believe that such practice disrupts the technical functioning of
> the DNS by breaking its specifications, thus impeding full
> interoperability between DNS clients and the registries which
> implement this practice.
> 
> We point out that such practice, if widely adopted, would cause
> significant confusion among the final users, give way to disputes
> about unfair competition, and, if further developed, leave to
> registries the opportunity to arbitrarily control the content of the
> network.
> 
> Thus we ask ICANN to ensure, by proper contractual agreements with
> registries, that all queries for domain names that have not been
> explicitly and singularly registered are answered with a message
> having an RCODE value of 3, as defined in RFC 1035, 4.1.1. Possible
> exceptions to this rule, if any, should be carefully limited and
> authorized in advance by explicit ICANN policy.
> 
> We also ask all ccTLD managers - including those not bound to ICANN by
> contractual agreements - not to adopt this practice.
> -- 
> vb.                  [Vittorio Bertola - vb [at] bertola.eu.org]<---
> -------------------> http://bertola.eu.org/ <-----------------------
>