Re: Call for agenda items for Minneapolis

[Eric Brunner-Williams in Portland Maine <brunner@nic-naa.net>]

Ed,

The WG has an interest, if only to prevent "end-runs" that the IESG may not
catch, in any draft that purports to be proximal to, or responsive to, the
requirements draft. The better interest is that the draft may be of general
interest to the implementors.

A long-winded "agree" to the point that "the WG is [not] limited to just
the core/standard documents".

The operation of a registry, in the abstract, isn't something that admits
"interoperation", unless the registy has more than one operator, either
concurrently or (proximally) consequetively. 

The policy of of a ccTLD delegation, in the abstract, isn't something that
admits "interoperation", unless delegation is multiply concurrently or
(proximally) consequetively instantiated.

In short, if we close our eyes to the fact that this is a specific operator
and a specific ccTLD, it is difficult to find where a second implementation
that could be interoperable, or not, is technically [1] feasible. If it makes
a difference that the delegation is ".US", or the operator favored as "second
after Verisign", then the WG is (temporarily) embarked on some national or
marketing detour.

We adopted XML. This is one of the distinguishing features of EPP.

There are alternatives:

	o RRP (versions 1.0 and version 1.1) uses name-value pairs,
	o SRS (versions 1.0, 1999 and version 2.0, 2001) uses name-value
	  pairs [2],
	o others, specifications not generally available

For reasons others can enquire into, the authors of this draft chose not to
use a schema, and opted out of XML as a design constraint. I noted earlier
that a name-value space similar to that proposed in this draft has been
investigated in the W3C PICS activity. It is hard to describe relationships
using this techniques [3], as most everyone involved, including the current
IAB chair, knows.

We could, without prejudice to the draft, consensus that name-value pairs
are out-of-scope for the PROVREG WG. I think we did this some (long) time
ago, but maybe not.

Well, I've got to go pick up kids.

Eric

[1] Any of these -- a partitioned .US registry (or a competitive registry),
multiple applicable policies for .US registries, registrars, and registrants,
and (to be sure) a non-unique delegation, would be interesting, and novel.
None is proposed.

[2] Elmar Knipp <Elmar.Knipp@knipp.de> December 10, 2001.

[3] http://www.w3.org/2001/Talks/09xx-semwai/all