Re: Comment on Fragment Identifiers

Arthur is correct.  Reducing the expressiveness as suggested will eliminate the possibility of message exchanges which have-or-may-have more than one message in a given direction.  I oppose such a change (it would be understandable to implement that variation in wsdl 1.1, however, given that it only defines two message exchange patterns (four, but two are marked "don't-use" in the spec)).

Amy!
On Wed, 13 Dec 2006 17:31:34 -0500
Arthur Ryman <ryman@ca.ibm.com> wrote:

>Ashok,
>
>This is a case of simplicity versus consistency with WSDL 2.0.
>
>In WSDL 2.0, the MEPs are an extension point and that third parameter
>can have any value (as defined by a new MEP). It's the message label
>and not restricted to in and out. It defines a role.
>
>I'm not advocating either way - just explaining the origin.
>
>Arthur Ryman,
>IBM Software Group, Rational Division
>
>blog: http://ryman.eclipsedevelopersjournal.com/
>phone: +1-905-413-3077, TL 969-3077
>assistant: +1-905-413-2411, TL 969-2411
>fax: +1-905-413-4920, TL 969-4920
>mobile: +1-416-939-5063, text: 4169395063@fido.ca
>
>
>
>"Ashok Malhotra" <ashok.malhotra@oracle.com> 
>Sent by: www-ws-desc-request@w3.org
>12/13/2006 05:08 PM
>
>To
>"www-ws-desc@w3.org." <www-ws-desc@w3.org>
>cc
>"public-ws-policy@w3.org" <public-ws-policy@w3.org>
>Subject
>RE: Comment on Fragment Identifiers
>
>
>
>
>
>
>
>Resending.  Last attempt was truncated.
>
>As you may know, the WS-Policy WG has been doing some work on defining
>element identifiers for WSDL 1.1 elements.  We are trying to align this
>work with the WSDL 2.0 fragment identifiers described in Appendix A.2
>of the WSDL 2.0 Candidate Recommendation draft of 2006-03-27.
>
>In looking at Appendix A.2 I came across two situations where I think
>the syntax can be improved.  Consider
> wsdl.interfaceMessageReference(interface/operation/message)
>this fragment identifier takes 3 parameters.  The first two take names
>as values while the third takes a message label whose value can only
>be "input" or "output".  Having a parameter that takes a keyword as
>value seems foreign to the general design in which parameters take
>names as values.  Thus, I suggest that the label be added to the name
>of the fragment identifier and it have only two parameters, thus:
>                 wsdl.interfaceMessageInput(interface/operation)
>                 wsdl.interfaceMessageOutput(interface/operation)
>
>The following row in the table can also be improved.
> wsdl.interfaceFaultReference(interface/operation/message/fault)
>can be replaced by two identifiers
>                 wsdl.interfaceInFault(interface/operation/fault)
>                 wsdl.interfaceInFault(interface/operation/fault)
>
>Similar suggestions apply to 
>                 wsdl.bindingMessageReference
> (binding/operation/message) 
>and
> wsdl.bindingFaultReference(binding/operation/message/fault)
>
>I hope you will consider these changes.
>
>All the best, Ashok
>
>
>
>


-- 
Amelia A. Lewis
Senior Architect
TIBCO/Extensibility, Inc.
alewis@tibco.com

Received on Thursday, 14 December 2006 16:14:06 UTC