Re: [dbound] comments on draft-deccio-domain-name-relationships-00

[Casey Deccio <[e-mail]>]

On Mon, Apr 6, 2015 at 6:33 PM, Kurt Andersen <[e-mail]> wrote:

>
> On Mon, Apr 6, 2015 at 3:28 PM, Casey Deccio <[e-mail]> wrote:
>
>> On Mon, Apr 6, 2015 at 6:02 PM, Andrew Sullivan <[e-mail]>
>> wrote:
>>
>>>
>>> I am not even a little bit convinced that the distinction is part of
>>> any solution.
>>
>>
>> Are you saying that there is no solution that can effectively use a
>> public/private distinction of names to determine whether or not there is
>> some sort of policy relationship between them?
>>
>
> I think that Andrew is saying, and I agree, that the names "public" and
> "private" are not useful any more in the dbound discussions. They are badly
> overloaded and, at best, vaguely defined.
>

I think you'll find that the terminology used in the draft being discussed
is consistent with descriptions within RFCs that use the terms (see RFC
6265 and RFC 7489, for example).

By overloaded, are you mostly referring to the "split" in the PSL that has
been noted by several individuals (i.e., the sections labeled "ICANN
DOMAINS" vs. "PRIVATE DOMAINS" in the file)?  Or do you mean something
else?  The draft makes an attempt to model the two sections of names--to
consider their conceptual merit.  But it (the draft) doesn't refer to them
in the same way the PSL does (i.e., by "ICANN DOMAINS" and "PRIVATE
DOMAINS").  Perhaps it could be more explicit in referencing those sections.

Among the solution considerations in Section 6 is the creation of a scope
definition resource (i.e., a PSL-like registry) to supplement a per-name
based solution.  While not explicitly stated in the draft, my personal
feeling is that the section of names that is under the section "PRIVATE
DOMAINS" (or "lower-level" public names, as denoted in the draft) should
*not* be candidates for such a registry.  Though that is not explicitly
noted in the draft.

Regarding "public" and "private" not being useful, I disagree.  See
Sections 3 and 6 in the -00 draft.

We would be better served to start referring to "foo" names and "bar" names.
>

I'm not sure that makes any sense.


> But I think that categorization in general is still a misdirection to the
> task of identifying names which share policy scopes.
>

There seems to be a lot of hang-up on the use of the terms "public" and
"private" in the draft.  I can't help but feel that some of this results
from an effort to distance ourselves from current "makeshift" notions and
solutions.  However, please don't throw out the baby with the bathwater.
While there is a general feeling that the existing solutions are misused
and inadequate, there is value in the concepts that they have introduced.
The draft at hand is not meant to be a solution, but rather presents a
picture of current notions for considerations for solutions, which current
notions include implementation (PSL) and specification (RFCs).
Understanding the starting point does not imply constraint by that starting
point.

That being said, the on-list discussions on this topic/draft can only be
contextually accurate if participants have read the -00 draft at hand, so
this is a general (i.e., not directed) request to read it if you haven't.

Regards,
Casey