+1,
Have the options of working with raw URN is a good relaxation that adds
a lot for thoes who work with constructed, generated URN where parts of
the URN indicates some or comes from semantical elements.
thanks
/anders
Farrukh Najmi wrote:
>
> Matt commented on regrep-rs-3.0-draft-01.pdf section 5.1.2 where it
> says "The id MUST be a Universally Unique Identifier (UUID) and MUST
> conform to the format of a URN that specifies a DCE 128 bit UUID as
> specified in [UUID]" as follows:
>
> "There are certain situations where it may be necessary for a registry
> to specify an ID That is not dce 128. For example, in Adobe's
> implementation, all user information comes from our Policy Server, and
> it has its own IDs. we use a namespace to indicate this, e.g.
> urn:aps:node:87787887878787. This section should be lenient to this."
>
> I have long felt that we should relax the id attribute to allow any
> URN and require that the registry check for uniqueness within the
> local registry during submission. In a federated situation the local
> id is implictly qualified by the home attribute value for the home
> registry so that would allow global uniqueness.
>
> Does anyone object to the change that allows RegistryObject.id to be
> any URN value that is unique within the local registry? If so please
> indicate your reasons clearly and indicate whether you could live with
> this change if others felt it was needed. Thanks.
>
