Re: [regrep-comresolve] Re: [regrep-comment] RIM 2.0 Comments/Questions

[Farrukh Najmi <Farrukh.Najmi@Sun.COM>]

Yes apparently both him and Mark Mahadeo are implementing a registry. This is all very exciting.

"Patil, Sanjaykumar" wrote:

> Look good to me.
>
> BTW, is this a new ebXML Registry implementer? I mean had we heard about them before? Exciting, isn't it.
>
> thanks,
> Sanjay Patil
> ----------------------------------------------------------------------------------------------------------
> IONA
> END 2 ANYWHERE
> Phone: 408 350 9619                                 http://www.iona.com
>
> -----Original Message-----
> From: Farrukh Najmi [mailto:Farrukh.Najmi@Sun.COM]
> Sent: Monday, April 08, 2002 9:44 AM
> To: 'Oasis RR Comment Resolution'
> Subject: [regrep-comresolve] Re: [regrep-comment] RIM 2.0
> Comments/Questions
>
> Team,
>
> Please review my proposed response to this email and send email. I would
> like to post a response soon.
> Thanks.
>
> Goran Zugic wrote:
>
> > We hope that you don't mind if we just add few comments/questions
> > belonging to the OASIS/ebXML Registry Information Model (RIM) v2.0
> > specification. We already implemented RIM v1.0 specification and
> > currently we are finishing an implementation of the RIM v2.0
> > specification.  We would like to share with you some of the issues
> > regarding the v2.0 specification:1) ServiceBinding and
> > SpecificationLink are missing Service and ServiceBinding references,
> > respectively. In our current implementation we added a Service
> > reference
> >    (service) to the ServiceBinding class and a ServiceBinding
> > reference (serviceBinding) to the SpecificationLink class.
>
> The ServiceBinding schema element is always composed within a parent
> Service element. Likewise the
> SpecificationLink schema element is always composed within a
> ServiceBinding element. Consequently
> they do not need (nor have) a reference to their parent object. This is
> deliberate and not a mistake.
>
> > 2) Section 9.6.2 on page 40:
> > "...
> > 9.6.2 Confirmation of Extramural AssociationsExtramural associations
> > may be thought of as a unilateral assertion that may not be viewed as
> > truth until it has been confirmed by the other (extramural) parties
> > involved (Users "u2" and "u3" in the example in section 9.5). To
> > confirm an extramural association, each of the extramural parties
> > (parties that own the source or target object but do not own the
> > Association) must submit an identical Association (clone Association)
> > as the Association they are intending to confirm using a
> > SubmitObjectsRequest. The clone Association must have the same id as
> > the original Association.
> > ..." The clone Association cannot have the same id as the original
> > Association because the id is a Universally Unique ID (UUID).  3)
>
> The cloned Association is the same Object as the original and therefor
> has the same id. There is one Association
> object with one UUID that is globally unique but has two clones. This is
> deliberate and not a mistake.
>
> > getOrganizations(String type) method in RegistryObject should be
> > changed to getOrganizations(LongName type).  4) Section 10.3.5 on page
> > 50:
>
> This is a minor error that we will fix in a future version of RIM We
> have logged in our issues list as:
>
> Issue 175: getOrganizations(String type) method in RegistryObject should
> be changed to getOrganizations(LongName type).
>
> Resolution: This minor error will be fixed in a future version of RIM.
> The information however is available to the reader by looking at the
> data type for the associationType attribute of class Association.
>
> >
> > "...
> > 10.3.5 Attribute nodeRepresentationIf the Classification instance
> > represents an external classification, then the nodeRepresentation
> > attribute is required. It is a representation of a taxonomy
> > element from a classification scheme. It is the responsibility of the
> > registry to distinguish between different types of nodeRepresentation,
> > like between the
> > classification scheme node code and the classification scheme node
> > canonical path. This allows client to transparently use different
> > syntaxes for nodeRepresentation.
>
> >
> > ..." Could you please give an example of the nodeRepresentation
> > attribute value that is "a representation of a taxonomy element from a
> > classification scheme."?Could you also please give an example of the
> > external
>
> For example consider the taxonomy is NAICS and the value is Book
> Publishers. In this case the
> nodeRepresentation is "51113" which is the NAICS code for "Book
> Publishers".
>
> > Classification getPath() and getCode() methods?
>
> getPath method may be used like an attribute in queries. See example in
> RS 2.0 at:
>
> 8.3.7.4 Getting Objects Classified By a ClassificationNode line 3455.
>
> The getCode method may be used like an attribute in Classification
> queries where the taxonomy value is needed
> without caring whether it is an internal Classification or an external
> Classification.
>
> You could have a query like:
>
> SELECT * from Classification where code LIKE "&Book%"
>
> > 5) Could you please explain how an owner of a registry object can be
> > determined from the RIM v2.0 security model?
>
> The owner of a RegistryObject can be determined by looking at the first
> AuditableEvent in the Audit trail
> for the RegistryObject and using its user attribute.
>
> > 6) We believe that v2.0 had to be backward compatible with V1.0. We
> > hope that you will take care of the backward compatibility issue in
> > the future specification versions/releases.
>
> v2.0 is not backward compatible with v1.0. We have firm plans for
> maintaining backward compatibility between
> v2.0 and v3.0.
>
> > Regards,
> > Duka Krunic & Goran Zugic
> > Members of ebXMLsoft Development Team
> > www.ebxmlsoft.com
>
> --
> Regards,
> Farrukh
>
> ----------------------------------------------------------------
> To subscribe or unsubscribe from this elist use the subscription
> manager: <http://lists.oasis-open.org/ob/adm.pl>

--
Regards,
Farrukh