W3C home > Mailing lists > Public > public-ws-chor@w3.org > March 2004

RE: pdf - WS Choreography Description Language, Version 1 (Edit or's Draft)

From: Liu, Kevin <kevin.liu@sap.com>
Date: Sat, 13 Mar 2004 02:53:20 +0100
Message-ID: <99CA63DD941EDC4EBA897048D9B0061DA95EDD@uspalx20a.pal.sap.corp>
To: public-ws-chor@w3.org
Cc: steve@enigmatec.net, martin.chapman@oracle.com, "'Greg Ritzinger'" <GRitzinger@novell.com>, "Trickovic, Ivana" <ivana.trickovic@sap.com>

To fulfill my action item, I am rephrasing my comments to the CDL draft so they can be recorded as issues:

>>Defition of Choreography and CDL
----------------------------------

The definition of CDL and the definition of Choreography in the requirement document should be synchronized. 

I like this one better than what's in the requirement document except one thing:

Why we are emphasizing "cross-enterprise" in the definition? Does a choreography have to be cross-enterprise? Can I define a choreography for departments within my enterprises?

I believe the key differentiator of WS-CDL and WSCI is the that CDL treats all participants equal whereas WSCI describe the interaction centered around ONE particular participant.

I would suggest we replace the word "cross-enterprise" with "peer-to-peer" here and in other places in this document except for cases where cross-enterprise collaboration is addressed

>>Precision of Specification 
------------------------------
In general, the draft reads more like a technical journal paper than a standard specification. Choreography defines interfaces among different participants, precision and un-ambiguity of the specification is ultimately important.  I feel it's really necessary to follow the WSDL2.0 model, and use a more formal mechanism such as XML infoset

>>Collaboration Types
---------------------
After reading section 2.1, I am still not clear why we need the "three" choreography types. What's the business motivation for such grouping? Is it really necessary to introduce some complexity and potential ambiguity in the spec level? For example, 
-	"abstract" vs "concrete" are quite different grouping criteria than "portable".  
-	 What's the difference between a "concrete" choreography and an "executable" process?

>>MEP support
-------------
Section 2.2, bullet for "interaction": Why only two MEPs are supported ? Please note WSDL2.0 defines a set of MEP and allow custom MEP to be defined for specific business needs

>>Definition of Choreograpy Transaction
---------------------------------------
Section 2.2 states "Choreography Transactions describes how to specify what additional Interactions should occur to reverse the effect of an earlier completed choreography." is it actually "composentation"? Using the term "transaction" here may be confusing to many people

>>Variable definition
---------------------
Section 2.4.2 is not clear about what kind of data a "variable" as defined here can contain. Should it be limited to info that are relevant to interaction? For example, in a credit approval choreography, while it's necessary to communicate the "approve"/"reject" decision,  I don't see any reason a choreography should expose data used to make such decision.  

In addtion, variable definition is another example that different grouping criteria are used. I am very lost in trying to understand  the differences of each kind of variables here

I also note that the definitions are the same for "information exchange variable" and "information variable".


Editorial issues
==================== 

>>Editors List:
Does this draft still only contain inputs from Oracle? I have been assuming that the WG already has an editing team aiming to converge inputs from Oracle, Commerce One, WSCI, requirement document, etc.

>> 4th paragraph of Introduction section
As an trivia example of my concern about "precision", when is "current"?
By the time this spec is published, the "current" status will be certainly changed. 

>>figure 1
 Does the group agree with the diagram? To me. It's at best controversial. For example, I see the following differently
-	Transactions etc are layers above RM
-	Security should  be vertically across all layers. 
-	How about manageability?
- How about potential layers above choreography?

It might be a good idea to include a revised diagram in the primer, but I don't feel comfortable to include it in the specification. Instead of defining our own, we should reference the architectural view from WSA  whenever possible 

>>section 1.4
- Title needs to be updated - not only WSDL is addressed here 
- Need updates: WSDL2.0.
-  Is XPath 2.0 already there?

>>Section 3. example
Does the choreography have to be centered around one role, in this case, the seller? 

For example, if the credit checking service needs to interact with a third party credit scoring service to get a FICO score. Can that interaction be added to this choreography?

>> WSDL example

This is not a  valid WSDL definition. 
- A WSDL 1.1 or 2.0 definition can only contain 0 or 1 <types> element.
- The xsd definitions have to be wrapped by a <schema> element under <types>
- There is no <message> element in WSDL2.0

>>Section 4
I strongly recommend we consolidate section 2 and 4.  I understand section 2 is intended to describe the language in a conceptual level, but in MANY places it's actually tied to particular language elements. Consolidate the "conceptual" info into relevant elements will at least make the spec more readable.

Hope this fulfill my AI.

Best Regards,
Kevin
 

-----Original Message-----
From: public-ws-chor-request@w3.org [mailto:public-ws-chor-request@w3.org] On Behalf Of Liu, Kevin
Sent: Tuesday, Mar 02, 2004 12:20 AM
To: public-ws-chor@w3.org
Cc: steve@enigmatec.net; martin.chapman@oracle.com; 'Nickolas Kavantzas'; 'Jeff Mischkinsky'
Subject: RE: pdf - WS Choreography Description Language, Version 1 (Edit or's Draft)


Hi Nick, Hi all,

I finally got a chance to read the draft again. Attached please find my comments.

I am certainly grateful to the great work Nick has put into this draft. As a deliverable of the working group, I feel we still need a lot of editing work to make the spec more precise. In general, the draft reads more like a technical journal paper than a standard specification.  Choreography defines interfaces among different participants, precision and un-ambiguity of the specification is ultimately important. I feel it's really necessary to follow the WSDL2.0 model and rewrite the spec using XML infoset.

Best Regards,
Kevin
 

-----Original Message-----
From: public-ws-chor-request@w3.org [mailto:public-ws-chor-request@w3.org] On Behalf Of Jeff Mischkinsky
Sent: Monday, Mar 01, 2004 06:11 AM
To: public-ws-chor@w3.org
Subject: Re: pdf - WS Choreography Description Language, Version 1 (Editor's Draft)



Hi,
   Here's the pdf version:

http://lists.w3.org/Archives/Public/www-archive/2004Mar/0002.html

cheers,
jeff

At 03:29 AM 3/1/2004, Greg Ritzinger wrote:

>The current thinking on the ws-choreography description language can be
>found at the following URL:
>
>http://lists.w3.org/Archives/Public/www-archive/2004Mar/0001.html
>
>Nick / Greg

Jeff Mischkinsky                      jeff.mischkinsky@oracle.com
Consulting Member Technical Staff     +1(650)506-1975
Director, Web Services Standards      500 Oracle Parkway M/S 4OP9
Oracle Corporation                    Redwood Shores, CA 94065
Received on Friday, 12 March 2004 20:54:08 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 18 December 2010 01:00:54 GMT