W3C home > Mailing lists > Public > public-ws-policy@w3.org > November 2006

RE: RE: Revised proposal for Bug 3730

From: Asir Vedamuthu <asirveda@microsoft.com>
Date: Tue, 31 Oct 2006 16:19:09 -0800
Message-ID: <1E0F0378382054439F14D5450650478F0B0D1699@RED-MSG-42.redmond.corp.microsoft.com>
To: Christopher B Ferris <chrisfer@us.ibm.com>, <public-ws-policy@w3.org>

> Why wsdlRef? Why not just uriRef or URIRef?

Sure - uriRef or URIRef - either works. We are indifferent on what the element name should be for such a domain expression. Ashok will not be attending tomorrow's call [1]. I suggest that we defer 3730 & 3599 until Ashok is here.

[1] http://lists.w3.org/Archives/Member/member-ws-policy/2006Oct/0021.html 

Regards,
 
Asir S Vedamuthu
Microsoft Corporation





From: public-ws-policy-request@w3.org [mailto:public-ws-policy-request@w3.org] On Behalf Of Christopher B Ferris
Sent: Monday, October 30, 2006 6:13 AM
To: public-ws-policy@w3.org
Subject: RE: RE: Revised proposal for Bug 3730


Why wsdlRef? Why not just uriRef or URIRef? The semantic of the interpretation of the 
fragment identifier (if present) is defined by the media type of the retrieved representation. 
Why would we have/need a different domain expression for another 
class of documents for which the means of identifying the subject was a URI reference? 

As for the question of namespace and whether there are more than one documents, these questions 
are orthogonal. 

Just because an element carries the same namespace does not necessarily mean that it is in a separate 
document. It means that it is in a separate namespace. Recall that the xml:id is in the same namespace as 
XML 1.0, yet it is clearly in a separate document. You can add elements and attributes to a namespace 
after the fact (although, clearly the xml:id example created some problems, but that was because another 
WG made the false assumption that a namespace would not be added to after-the-fact). 

Let's all remember that a namespace is a "space of names". Nothing more, nothing less. 

I frankly do not see a compelling reason to introduce a separate namespace for this. 

As to use of a separate document, I also fail to understand the motivation. It isn't as if the 
semantic of dereferencing a URI is all that novel. Are we really concerned that this cannot 
be implemented? 

Cheers, 

Christopher Ferris
STSM, Software Group Standards Strategy
email: chrisfer@us.ibm.com
blog: http://www.ibm.com/developerworks/blogs/page/chrisferris
phone: +1 508 377 9295 

public-ws-policy-request@w3.org wrote on 10/27/2006 08:30:09 PM:

> 
> Thank you Ashok!
> 
> >I propose we call the wrapper element wsdlRef
> 
> +1
> 
> >b) A separate namespace for this element -- YES
> >c) A separate document for the domain expressions -- YES
> 
> Sounds good.
> 
> What do others think about questions a)-c)?
> 
> Regards,
>  
> Asir S Vedamuthu
> Microsoft Corporation
> 
> 
> -----Original Message-----
> From: Ashok Malhotra [mailto:ashok.malhotra@oracle.com] 
> Sent: Wednesday, October 25, 2006 1:06 PM
> To: Paul Cotton; Asir Vedamuthu; Daniel Roth; public-ws-policy@w3.org
> Subject: RE: RE: Revised proposal for Bug 3730
> 
> Here are my answers to the three outstanding issues:
> 
> a) One wrapper or two wrappers.
> The WG agreed on one wrapper.
> I propose we call the wrapper element wsdlRef used as below
> 
>    <wsp:AppliesTo>
>            <xxx:wsdlRef>
>  
> http://example.com.LoanFlow#wsdl.service(LoanFlowService) 
>            </xxx:wsdlRef>
>       </wsp:AppliesTo> 
> 
> The wrapping element xxx:wsdlRef is defined with the following XML
> Schema fragment.
> 
> <xs:element name="wsdlRef" type="anyURI"/>
> 
> b) A separate namespace for this element -- YES
> 
> c) A separate document for the domain expressions -- YES
> 
> All the best, Ashok
>  
> 
> > -----Original Message-----
> > From: public-ws-policy-request@w3.org 
> > [mailto:public-ws-policy-request@w3.org] On Behalf Of Paul Cotton
> > Sent: Wednesday, October 25, 2006 10:08 AM
> > To: Asir Vedamuthu; Ashok Malhotra; Daniel Roth; 
> > public-ws-policy@w3.org
> > Subject: RE: RE: Revised proposal for Bug 3730
> > 
> > 
> > We need a concrete proposal in order to close out the WSDL 
> > external attachment issues (issue 3730 and 3599).
> > 
> > The Oct 25 WG meeting touched on this thread and agreed that 
> > getting the following three questions answered via email 
> > would help us develop a concrete proposal for these two 
> > issues before the F2F meeting.
> > 
> > > (a) Two distinct wrapper elements for WSDL 11 and 20 sound good.
> > 
> > WG has expressed a preference for ONE wrapper element.  The 
> > MIME type of the resource should determine the semantics of 
> > the URI.  Asir and others expressed support for this position 
> > at the Oct 25 meeting.
> > 
> > > (b) These two wrapper elements (WSDL 11 and 20) are domain 
> > expressions.
> > > Just like other domain related work (security, reliability, 
> > > transaction, etc.), these domain expressions can be in their own 
> > > namespace names (for instance 
> > > http://www.w3.org/2006/07/ws-policy/wsdl20). By minting new 
> > namespace 
> > > names for domain expressions, WG can showcase this 
> > extensibility point.
> > 
> > WG has not yet decided if they want a separate namespace for 
> > this domain expression.
> > 
> > Pro position:
> > - Separates the "domain expression" into a separate namespace 
> > and keeps it separate from the main Policy namespace.
> > 
> > Con position:
> > - This might cause people to infer that the material in the 
> > second namespace is an "optional" feature that they don't 
> > have to implement in order to do WS-Policy.
> > 
> > > (c) Just like the policy language and assertions, the external 
> > > attachment mechanism and domain expressions evolve independently. 
> > > There is a clean separation between the external attachment 
> > mechanism 
> > > and domain expression. To promote these, we suggest 
> > documenting these 
> > > domain expressions in a separate document (for instance 
> > 'Web Services 
> > > Policy - Domain Expressions').
> > 
> > WG has not yet decided if they want a separate document.
> > 
> > Pro:
> > - This would permit the work to advance on its own schedule 
> > separate from the Framework and Attachment specs.
> > 
> > Please express your opinions on items b) and c) above via 
> > email before the Nov 1 distributed meeting.
> > 
> > /paulc
> > 
> > Paul Cotton, Microsoft Canada
> > 17 Eleanor Drive, Ottawa, Ontario K2E 6A3
> > Tel: (613) 225-5445 Fax: (425) 936-7329
> > mailto:Paul.Cotton@microsoft.com
> > 
> > 
> > 
> > 
> > 
> > > -----Original Message-----
> > > From: public-ws-policy-request@w3.org [mailto:public-ws-policy- 
> > > request@w3.org] On Behalf Of Asir Vedamuthu
> > > Sent: October 18, 2006 11:28 AM
> > > To: Ashok Malhotra; Daniel Roth; public-ws-policy@w3.org
> > > Subject: RE: Revised proposal for Bug 3730
> > >
> > >
> > > Thank you Ashok. There are three other bits on 3730 and 3599:
> > >
> > > (a) Two distinct wrapper elements for WSDL 11 and 20 sound good.
> > >
> > > (b) These two wrapper elements (WSDL 11 and 20) are domain 
> > expressions.
> > > Just like other domain related work (security, reliability, 
> > > transaction, etc.), these domain expressions can be in their own 
> > > namespace names (for instance 
> > > http://www.w3.org/2006/07/ws-policy/wsdl20). By minting new 
> > namespace 
> > > names for domain expressions, WG can showcase this 
> > extensibility point.
> > >
> > > (c) Just like the policy language and assertions, the external 
> > > attachment mechanism and domain expressions evolve independently. 
> > > There is a clean separation between the external attachment 
> > mechanism 
> > > and domain expression. To promote these, we suggest 
> > documenting these 
> > > domain expressions in a separate document (for instance 
> > 'Web Services 
> > > Policy - Domain Expressions').
> > >
> > > Regards,
> > >
> > > Asir S Vedamuthu
> > > Microsoft Corporation
> > >
> > >
> > > -----Original Message-----
> > > From: public-ws-policy-request@w3.org
> > > [mailto:public-ws-policy-request@w3.org] On Behalf Of Ashok Malhotra
> > > Sent: Tuesday, October 17, 2006 3:28 PM
> > > To: Daniel Roth; public-ws-policy@w3.org
> > > Subject: RE: Revised proposal for Bug 3730
> > >
> > >
> > > Dan:
> > > You said ...
> > > > ... but it doesn't say what the implied Policy Scopes and Policy 
> > > > Subjects are.  I'm guessing they should be the same as the Policy 
> > > > Scopes and Subjects already defined for WSDL 2.0 attachment.
> > >
> > > This is a good point.  If I look at the latest Policy 
> > Attachment spec 
> > > it has a long section on WSDL 2.0 attachment with an extensive 
> > > discussion of Policy Subjects and Scopes and merging.  I 
> > don't think 
> > > we want to repeat this material for the external attachment 
> > section.  
> > > So, how about we add a paragraph that says something like:
> > >
> > > The semantics of associating policies with WSDL 2.0 
> > components using 
> > > the external attachment mechanism are exactly the same as if the 
> > > policies had been attached directly to WSDL 2.0 components 
> > using the 
> > > mechanisms described in section 5. The possible policy scopes are 
> > > exactly those allowed in section 5.2 and the calculation of 
> > effective 
> > > policies is done in exactly the same manner as described in section 
> > > 5.4.
> > >
> > > All the best, Ashok
> > >
> > >
> > > > -----Original Message-----
> > > > From: Daniel Roth [mailto:Daniel.Roth@microsoft.com]
> > > > Sent: Wednesday, October 11, 2006 8:59 AM
> > > > To: Ashok Malhotra; public-ws-policy@w3.org
> > > > Subject: RE: Revised proposal for Bug 3730
> > > >
> > > > Hi Ashok,
> > > >
> > > > Thanks for sending out a new proposal.  I see that you added a 
> > > > wrapper element, which is great.
> > > >
> > > > The policy attachment spec uses the terms Policy Scope and Policy 
> > > > Subject when defining attachment mechanisms.  You attach 
> > policies to 
> > > > Policy Scopes which associates the attached policy with 
> > all Policy 
> > > > Subjects within that Policy Scope.  This language is also used to 
> > > > describe how policies are merged when multiple policies 
> > are attached 
> > > > to different scopes containing the same Policy Subject.  This 
> > > > proposal describes how to use a WSDL 2.0 component reference as a 
> > > > domain expression, but it doesn't say what the implied 
> > Policy Scopes 
> > > > and Policy Subjects are.  I'm guessing they should be the same as 
> > > > the Policy Scopes and Subjects already defined for WSDL 2.0 
> > > > attachment.
> > > >
> > > > Also the current proposal doesn't use any RFC language, and it 
> > > > probably needs to if you want the proposal simply copied into the 
> > > > attachment spec.  This could just be an action item for 
> > the editors.
> > > >
> > > > Daniel Roth
> > > >
> > > > -----Original Message-----
> > > > From: public-ws-policy-request@w3.org 
> > > > [mailto:public-ws-policy-request@w3.org] On Behalf Of 
> > Ashok Malhotra
> > > > Sent: Friday, October 06, 2006 2:15 PM
> > > > To: public-ws-policy@w3.org
> > > > Subject: Revised proposal for Bug 3730
> > > >
> > > >
> > > > As requested, I have defined an element wrapper for the URI 
> > > > Reference that indicated the WSDL 2.0 component.
> > > >
> > > > All the best, Ashok
> > > >
> > > >
> > >
> > >
> > 
> > 
> > 
> > 
> 
> 
Received on Wednesday, 1 November 2006 00:20:34 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:20:43 GMT