On Mon, Mar 3, 2014 at 1:20 PM, Joe Abley <[email protected]> wrote:
>
> On 3 Mar 2014, at 13:12, Ted Lemon <[email protected]> wrote:
>
>> On Mar 3, 2014, at 1:09 PM, Joe Abley <[email protected]> wrote:
>>> I suggest that it's entirely plausible for someone to choose a 
>>> DNS-namespace anchor for their non-DNS namespace that is as stable as they 
>>> want, depending on their needs.
>>
>> This is clearly not the case for an open protocol spec, though, since there 
>> is no one entity that could be responsible for maintaining the registration.
>
> Certainly there could be cases where that is true. I can't think of an 
> example from the candidates we've identified in this (and other threads) to 
> date, though.
>
> (e.g. tor -> eff.org; dns -> okturtles.com).
>

because I don't want my leaked query for www.nakedfurries.foo to hit
the wimble.example.com nameservers? Even if I currently trust them not
to be logging that.

Putting .foo under .alt and making it a locally served zone (return
NXDOMAIN for all queries) means that my leaked queries only hit my
local resursive. And not everyone's leaked queries get aggregated
somewhere. The ALT doc also suggests that stubs could drop queries,
and then they wouldn't even hit the recursive ( I suspect that this is
not likely, but...)

W



> There's inevitably *someone* who needs to take a lead on coordinating any 
> codebase, if the codebase is to produce anything useful. I'm suggesting that 
> that person could easily register a domain.
>
>
> Joe
> _______________________________________________
> DNSOP mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/dnsop

_______________________________________________
DNSOP mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dnsop

Reply via email to