RE: [ebxml-bp] [ebBP] 6/2/2004: [RSD] Comment on WI-43-66 Voting Item

["Steve Capell" <steve.capell@redwahoo.com>]

Monica,

Thanks for your response.  One point I strongly disagree with:  <recognizing
that the UUID would be assigned by the registry (not the business process
itself)>.  If the registry does that then we face severe interoperability
problems between federated registries - as well as human "unfriendly" IDs.
Identifiers for registry objects should be URIs that are derived from the
model and so should be publisher assigned and should be the same,
irrespective of which registry node they are published to.   I think I have
said before that, despite my liking of ebXML regrep in general, there is one
thing that the uddi group have done better and that is publisher assigned
keys and the key partition framework.

Regards,

Steve Capell
Red Wahoo Pty Ltd
+61 410 437854

-----Original Message-----
From: Monica J. Martin [mailto:Monica.Martin@Sun.COM] 
Sent: Thursday, 3 June 2004 8:21 AM
To: ebXML BP; Duane Nickull; Farrukh Najmi
Subject: [ebxml-bp] [ebBP] 6/2/2004: [RSD] Comment on WI-43-66 Voting Item

Discussion|OASIS.ebBP.WI43-66-Name and Name ID Clarification;
Topic|;
Point|Use of UUID;
Attachment|http://www.oasis-open.org/archives/ebxml-bp/200406/msg00000.html;



mm1@
Steve,
Thanks for the input and contribution.  As for the vote, all members of 
the TC are eligible to vote here (observers or not), so please proceed.

We did discuss the use of UUID and the xsd:id.  The latter could be 
perceived as an external identifier specific to the parties and their 
interactions.  That does not preclude the creation of a UUID for 
registry purposes. The Reg/Rep team, and the CC/Reg Review subteam, have 
been discussing that both may be important, recognizing that the UUID 
would be assigned by the registry (not the business process itself).

I've cc: Farrukh Najmi and Duane Nickull on this email so they are aware 
of the discussion (as they have been involved in past communication). In 
order for ebBP to support the <documentation>  attribute you specified, 
we would have to establish a variable that could acquire the UUID when 
the process definition is stored
(which is may or may not be).

I'll open the discussion to the team to consider your suggestion and the 
other associated criteria.

Thanks.
@mm1