[regrep] Distinguishing a Replica from an original RegistryObject

[Monica Martin <mmartin@certivo.net>]

See attached data point on this subject from KIEC (Jasmine Jang).

Thanks.
Monica
========================================================================
=================================================
From Farrukh:
>Thanks Monica. This is a good data point.

From Monica:
>>Farrukh,
>>Here is the helpful response I received from Jasmine Jang, KIEC, on
>>replication of Registry Objects.  This is an information point
>>validating our effort.
>>
>>Thanks.
>>Monica
>>

From Jasmine Jang, KIEC:
>>Monica,
>>
>>I've reviewed v.3.0 draft. Techical considerations are enough to build
>>the cooperations among various registries.
>>Thank you, Najmi, Stojanovic and MacKenzie so much.
>>In Korea, the regrep(v.2) are now implemented in Steel and Iron
>>Industry. Next, I have a plan to build the cooperation between KIEC's
and Steel's.

>>Be happiest, today.
>>
>>Best regards
>>Jasmine

From Monica:
>>Jasmine,
>>As the Reg/Rep team's works to move to a v3.0 specification, we are
>>discussing replicas for an original RegistryObject.  Looking at the
>open questions below, I believe that KIEC could provide some real-world
>input here so that what we propose and move forward with will meet
business
>>needs.
>>
>>Can you provide some insight or suggestions?  We can provide further
>>detail if needed.  I am basing this inquiry to you on replication
being
>>discussed in the brief you sent me in May 2002 (see attached).
>>
>>Thanks.

From Farrukh N:
>>The RS 2.2 spec introduces a new Object Replication feature.  Object 
>>replication allows a registry to create local replicas of 
>>RegistryObjects in other registries. The specification of this feature

>>raises a few questions:
>>
>>1. Should there be a way for a registry client to be able to tell 
>>whether a RegistryObject is an original or whether it is a replica?
Use
>>cases would be appreciated.
>>
>>Assuming that the answer to (1) is yes then it brings up additional 
>>questions on how the information model should distinguish a replica
>from an original.
>>
>>a) One possibility is that every RegistryObject have a boolean
>attribute isReplica
>>b) Another possibility is that every RegistryObject have a String 
>>attribute called home, which contains the baseURI to the home registry

>>for an object. It can be null if the home registry is the local
>>registry.

>>Both a) and b) are not desirable since they can have conflicting
>values.

>>Thanks in advance for your input and for getting input from KIEC.
-- 
Regards,
Farrukh