Re: host transfers

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

> This points to an issue that Scott Hollenbeck and Eric
> Brunner-Williams had words about a few days back.  Since it's come up
> twice in the last week, I'll throw the question(s) out there:
> 
> Does EPP need a registry-to-registry communication mechanism?
> If not, does a registry-to-registry communication mechanism need to
> exist separate from EPP?             

The first technical issue is can a registry initiate an instance of
communication. If so, then in the transfer object use case, the registry
can "push" state to the non-initiating registry, (or to a 3rd-party, as
ICANN gTLD regisry contracts call for a form of 3rd-party archival data
flow, and possibly to a distinct registry).

As a specification issue, and as a working group process and document
issue, we have a document that specifies an asymetric event model, aka
"client-server". Where does this constraint arise?

I hope we can remove the "client-server" bits from the abstract of the
current grrq, and from the tcp mapping doc, which are the only places
I know it exists (corrections welcome), and add a push command. I do have
the text ;-)

Once we answer the question "does EPP need a registry-to-anything"
communication mechanism (and our experience is an unqualified "yes",
where "anything" == registrars, and use case == xfr, as preregproto
participants know, and we put it into our -00 XRP draft), then we can 
sensibly ask "registrars?, escrow-agents?, registries?"

Scott's proposed IM'ing as an OOB downstream notice mechanism. It isn't my
own first choice.

Eric