Re: [apps-discuss] "finding registered domains"

["J. Trent Adams" <[e-mail]>]

Andrew -

On 3/12/13 2:50 PM, Andrew Sullivan wrote:
> On Tue, Mar 12, 2013 at 04:37:10PM -0400, Murray S. Kucherawy wrote:
>
>> I am probably making some assumptions here, about which I should be careful
>> given the weird hat that's around here somewhere.  It's based on a property
>> of WHOIS that exists now, at least with numbers: if I ask for 1.2.3.4, then
>> I will get back the registration for the tightest containing subnet
>> registration, and possibly all of them upwards.  If a WEIRDS server did the
>> same with names (or could if asked), then example.com would be the returned
>> result for foo.bar.example.com.
> But why should you believe what example.com says about
> foo.bar.example.com?  If we could believe that, we wouldn't need a new
> mechanism.

There are times in practice when it's useful to make certain assumptions
like "if there is no response to a given query at a FQDN, hop all the
way to the registered domain to see what it says."  That's not to say
that this model fits all use cases, just that sometimes it's helpful
(and your proposal seems to support it by pointing me in the right
direction).

So, for example, I'd publish a SOPA record at "way.down.the.tree.org"
pointing to "tree.org".  Then if what I'm looking for doesn't exist at
"way.down.the.tree.org", I would know that the answer is at "tree.org".

This may actually be a bastardization of what you're ultimately trying
to accomplish, but does seem to serve my goals with the added benefit of
weening myself off of the public suffix list.

- Trent

>
> A
>

-- 
J. Trent Adams

Profile: http://www.mediaslate.org/jtrentadams/
LinkedIN: http://www.linkedin.com/in/jtrentadams
Twitter: http://twitter.com/jtrentadams