We had lots of discussions on the list since the Vancouver F2F, with mixed 
results. The chairs feel that it is time to move forward one way or the other.
We think the best way forward is to appoint an editor who will merge the 
non-controversial parts of the currently available two individual submissions, 
draft-das and draft-wei, including the agreements we have made so far on the 
list, with the intent to create a document for wg adoption. Vincent Chen has 
volunteered to take the very rewarding job of editor.
Here's a summary of what we have agreed so far and what is non-controversial 
and should be included into the merged document:
Intro, protocol description, protocol functionalities
There was no objection to have a DB initialization message, no controversy 
around the registration/ validation/query/reporting

We also agreed that the Discovery process should result in discovering both the 
URI of the DB and the regulatory domain. We have not agreed whether we should 
include the discovery as part of this document or have a separate document. We 
have also not agreed what mechanisms to use for discovery, LoST or the one in 
draft-Probasco. In my opinion, even if we use LoST, we'd need a document 
describing how to do LoST server discovery and define a parameter to be used 
for conveying the regulatory domain. Ie, either way, we need a document which 
describes these details (whether that document would stay as a standalone or 
rolled into the base solution doc is a separate discussion we could have). Any 
volunteer to generate such a document for the wg could discuss?
I propose the merged document to have a section on discovery and list at least 
the above assumptions.

We have not agreed on the encoding of the data elements. The chairs and the AD 
will soon come up with a process to help us choose between JSON and XML, unless 
we'll find good reasons to support and define both. For the time being, I'll 
ask the editor to at least put together from the two documents a list of 
parameters which we'll need to include into the protocol messages, noting the 
little agreements we made, like using ISO8601 format.

Regarding security, the proposals are to have the protocol be defined to handle 
both shared secrets and client certificates. Brian had a problem with the 
shared secret, saying that if we choose to support shared secrets but do not 
define a provisioning mechanism, the iesg may not like it. Brian has an AP to 
clarify this with the security experts within the iesg. Until this point is 
clarified, the editor should include the certificate part and have a 
placeholder for the shared secret part.

Let's start with the above. I am hoping to have a merged document produced by 
the editor in a relatively short period of time.


-          Gabor




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

Reply via email to