Re: [DNSOP] draft-liman-tld-names-04

[Eric Brunner-Williams <[e-mail]>]

On 11/24/10 9:23 AM, Andrew Sullivan wrote:
> On Wed, Nov 24, 2010 at 01:15:23PM +1100, James Mitchell wrote:
>> If deployed software does not work with a TLD, it is the TLD owner who loses.
>
> I'm sorry, but that claim is arrant nonsense.  We _all_ lose. ...

-1.

first, the likelihood of the string review failing to detect a 
"deployed software does not work with string" condition that is not an 
edge case, is not large, and the collection of evaluation processes 
documented [1] clearly anticipates failing a non-zero number of 
applications for non-technical, as well as technical reasons, some of 
which are difficult to find consensus as to the necessity and utility, 
e.g., v6 and dnssec as mandatory to implement, flavor and amount of 
continuity funding instrument, etc.

second, as paf observed, where the string is proposed as a label in 
any other context, e.g., as a non-root terminating component of a 
multi-label sequence, it is the parties with an interest in that name 
to address mapping, the parties with an interest in that specific 
resource, and no other party attempting any other possible act of 
resolution, possibly to the same resource (see the clone warz, in 
dnsext and elsewhere) that "loses".

less absolute claims would be helpful.

i was under-amused by the earlier "it is better for deployed software 
to break than ..." misstating of a reasonable engineering view on the 
cost of correctness and the "i resume you appreciate that ..." form of 
issue advocacy. that passes for acceptable social discourse on icann 
policy lists, and a whole lot worse. it is infra dig here.

-e

[1] http://www.icann.org/en/announcements/announcement-2-12nov10-en.htm