OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

regrep message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]


Subject: [regrep] Re: Core Components Specifications




> Message text written by Sandy Klausner
> >I've heard
> arguments for both UML and XML, both of which do not appear to adequately
> address the....
>>>>>>>>>>
STOP!!!!  THEY ARE TWO DIFFERENT EXPRESSIONS OF THE SAME THING!!!!!

Please at least acknowledge this.  Others see this.  Continuing along
this line of thinking is counter productive.

Farrukh wrote:

>Gentle folks,
>I cannot believe that we are debating between UML and XML.
>XML is a syntax while UML is a higher level thing to define an Object Oriented
>Model.  UML can and has used XML to express its model.
>>>>>>>>>>
Words of wisedom!

The question that needs to be answered is still posed:

"When you resolve a Registry reference to a Core Component, what do you
find?"

I made a list of the requirements that Architecture has for this
artifact in a previous thread.  

This question HAS to be answered before the Core Components can be
implemented.  If they are not implemented,  they do not become a
standard under UN/CEACT rules.  Therefore,  I humbly request that we
have some sensible debate over this and arrive at an answer.

The ideas hinted at by Matt and Farrukh are very good. However, I would
like to hear an *official* answer from the CC team.  We need the answer
before we can implement anything.

Original message is reposted below:

*******Request for Comments*******

The ebXML Registry contains an <ObjectReg> element in the Regsitry XML
Schema which points at the RegistryObject.

What is the <ObjectRef> Element going to point at when it references
a Core Component?

I aks this question becuase this needs to be answered BEFORE core
components can be implemented.  <QuantumJumpForwardInLogic> Since the
UN/CEFACT approval process calls for two independant implementations, 
this question must be answered and in the CC Specification.
</QuantumJumpForwardInLogic>.

To solve this question, I advocate that we first define all the
requirements for the answer.  Here is a starter list compiled from the
comments on this thread so far (from a programmatic AND business level):

1. Whatever the Registry references must not be bound to only being
expressed in one single syntax however, it MUST be expressed in at least
ONE syntax that is normalized. Otherwise, implementors will not know how
to parse the Core Component to present representations fo it to other
system actors, whether they be machine or human.

2. It should probably contain easily recognizable, traversable links to
the UML model representation of the Core Component AND have expandable
linking abilities to be capable of displaying itself in other formats
(in case something eventually supercedes UML, XML et al as per Lisa S's
comments). I would advocate that the linking mechanism be standardized.

3. It should be parsable by applications in a standardized manner.  This
is to avoid data serialization errors like misinterpretting a signed
integer as an unsigned interger.

4.  It should be easy to transform it via an application to other syntax
specific representations of the same object for re-use.

5.  It should ideally be either human parsable (readable) or it should
be easy to build an application to present a human being a link to a
human understandable representation of the Core Component (ie - text or
UML). 


These represent a good start to the requirements for this yet to be
determined artifact.

Duane Nickull
-- 
CTO, XML Global Technologies
****************************
Transformation - http://www.xmlglobal.com/prod/foundation/
ebXML Central - http://www.xmlglobal.com/prod/central/


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]


Powered by eList eXpress LLC