Re: [ebxml-cppa] CPPA Version 1.05

[Christopher Ferris <chris.ferris@sun.com>]

Dale,

Some comments below.

Cheers,

Chris

Dale Moberg wrote:

> Chris,
> It seems to me that some sort of URI involving
> a UDDI registry could indeed be sufficient
> for the PartyRef value. That would
> be one extension. An ebXML registry might be another.


Right, the point I was making w/r/t UDDI is that UDDI
provides a definition for a business already. Whether it
includes all of the requisite bits of information that
one might ever conceive is another thing, but it is certainly
a start.


> The current dilemma is how best or, more accurately,
> just _how_ to handle this sort of 
> extensibility while having some chance at automation
> (or interoperable operation) and while having
> a specification that is frozen with respect to
> the enumerated values it contains. (Woops, a couple
> of current dilemmas.) So, we are hoping that we can
> get Oasis to set up something like an IANA for
> URNs that identify enumerated values, and let this
> be the official site for finding new enumerated
> values, beyond what gets documented in specification


Well, that's what a registry is for:) Seems to me that one
could use either UDDI or ebXML reg/rep to satisfy the
need to "register" identifiers for enumerated values...


> version X.X. These enumerated values then are possible
> ways of filling in the attributes "type=XXX" that we
> seem to be adopting to allow some semantic constraint
> on the values (as under the PartyRef element and the
> PartyId element and elsewhere.) Duane's concern has
> been to have better support for automation and I
> agree with him that we need to try to arrive at a 
> better system.


I agree that we need to be working towards better automation.
However, I would hope that this were done in a manner that
did not *mandate* a specific approach.


> 
> So, do you have a proposal for how to handle the
> type specifier for UDDI? Words of explanation of
> how to retrieve the right value? Send them up by
> next Monday! 


Well, the UDDI business_entity (I think that's the name)
is defined in a namespace qualified schema. The "type"
could simply be the namespace identifier, or possibly
a qualified name (QName).

e.g.

	<eb:PartyRef xmlns:uddi="..." eb:type="uddi:business_entity" .../>

This might be better than just a URI for the type, as it provides
a lot more information and can be scoped.


> 
> Dale
> 
> -----Original Message-----
> From: Christopher Ferris [mailto:chris.ferris@sun.com]
> Sent: Friday, January 25, 2002 9:33 PM
> To: Tony Weida
> Cc: Duane Nickull; CPPA
> Subject: Re: [ebxml-cppa] CPPA Version 1.05
> 
> 
> Just out of curiosity, could someone explain to me why
> say a UDDI reference to a business_entity wouldn't be
> sufficient for purposes of the PartyRef?
> 
> Seems to me that this might be a reasonable choice for
> some and could garner some additional good will w/r/t
> ebxml and its relationship to the de facto web services
> components.
> 
> Just a thought.
> 
> Cheers,
> 
> Chris
> 
> Tony Weida wrote:
> 
> 
>>Duane,
>>
>>My responses are inline ...
>>
>>Tony
>>
>>----- Original Message -----
>>From: "Duane Nickull" <duane@xmlglobal.com>
>>To: "Tony Weida" <rweida@hotmail.com>
>>Cc: "CPPA" <ebxml-cppa@lists.oasis-open.org>
>>Sent: Friday, January 25, 2002 4:39 PM
>>Subject: Re: [ebxml-cppa] CPPA Version 1.05
>>
>>
>>
>>
>>>>Tony Weida wrote:
>>>>
>>>>CPPA version 1.05 is attached.  The zip file includes the
>>>>specification, the XSD, an example CPA, and example CPPs.  This is
>>>>the last version before the F2F.
>>>>
>>>>
>>>Tony et al:
>>>
>>>Will the comments regarding the link to a recognizable XML format for
>>>the Party details be addressed at the face to face? (ie - necessary
>>>
> for
> 
>>>context to be implemented).  HTML is NOT acceptable for this.  What
>>>happens if I were to write my party information in HTML using
>>>
> characters
> 
>>>that you can;t read (ie - Japanese Kanji, Korean, Hebrew etc.).
>>>
>>>
>>Issue 151, Specify type for PartyRef, was targeted for consideration
>>
> during
> 
>>the version 1.1 time frame and will be discussed at the F2F.
>>
>>A CPP author can already identify a type according to the current
>>
> spec.
> 
>>If the type is not identified, then by default (and only by default)
>>
> the
> 
>>referenced document must be HTML.
>>
>>Personal opinion: if a CPP author chooses to use HTML and some natural
>>language that I don't understand, that IS acceptable -- even if you or
>>
> I
> 
>>might feel that "better" choices are available.  Anyone who finds it
>>unacceptable can, of course, take their business elsewhere.
>>
>>
>>
>>>Also,  what is the status of the oid:urn investigation?  I know that
>>>Dale M. is looking into that with OASIS however, using "tp:type="DUNS"
>>>is currently a totally unnacceptable alternative since Dun +
>>>
> Bradstreet
> 
>>>numbers are not given out globally or easily nor has D+B set up a
>>>
> lookup
> 
>>>Registry.
>>>
>>>
>>I believe that discussions with OASIS are ongoing.
>>
>>
>>
>>>I humbly suggest that these two items still be treated with the
>>>
> highest
> 
>>>priority. The current v 1.05 still has not addressed these.  Is this
>>>discussion planned for the F2F?
>>>
>>>Duane Nickull
>>>--
>>>CTO, XML Global Technologies
>>>****************************
>>>Transformation - http://www.xmlglobal.com/prod/foundation/
>>>ebXML Central - http://www.xmlglobal.com/prod/central/
>>>
>>>
>>>
>>----------------------------------------------------------------
>>To subscribe or unsubscribe from this elist use the subscription
>>manager: <http://lists.oasis-open.org/ob/adm.pl>
>>
> 
> 
> 
> ----------------------------------------------------------------
> To subscribe or unsubscribe from this elist use the subscription
> manager: <http://lists.oasis-open.org/ob/adm.pl>
>