Web Services Addressing Teleconference

16 Apr 2007

See also: IRC log


Anish Karmarkar (Oracle Corporation)
David Hull (TIBCO Software, Inc.)
David Illsley (IBM Corporation)
Gilbert Pilz (BEA Systems, Inc.)
Mark Little (JBoss Inc.)
Ram Jeyaraman (Microsoft Corporation)
Robert Freund (Hitachi, Ltd.)
Tom Rutt (Fujitsu Limited)
Yin-Leng Husband (HP)
Amelia Lewis (TIBCO Software, Inc.)
David Orchard (BEA Systems, Inc.)
Eisaku Nishiyama (Hitachi, Ltd.)
Francisco Curbera (IBM Corporation)
Glen Daniels (Sonic Software)
Jacques Durand (Fujitsu Limited)
Katy Warr (IBM Corporation)
Marc Goodner (Microsoft Corporation)
Paul Knight (Nortel Networks)
Philippe Le Hégaret (W3C)
Prasad Yendluri (webMethods, Inc.)
Rama Pulavarthi (Sun Microsystems, Inc.)
Tony Rogers (Computer Associates)
Yves Lafon (W3C)
Bob Freund
Bob Freund




<scribe> scribe: Bob Freund

resolution: minutes of 2007-04-02 accepted

New issue http://lists.w3.org/Archives/Public/public-ws-addressing-comments/2007Apr/0000.html

resolution: accepted as proposed

New issue http://lists.w3.org/Archives/Public/public-ws-addressing/2007Apr/0003.html

by Tom R wrt attaching a policy to an epr

TomR: Addr did the wsdl, why not the policy?
... concern is that there will be a void if left unanswered

Gil: I don't see how we can be tasked on how to put policy in an epr.

<TRutt_> WS addressing has decided how to attach wsdl to the EPR, we could do otherwise

Ram: MEX seems to be developing as a way to get this done

Tom: this method is not yet on a standards track

Anish: Why would MEX stop a wg from defining the functionality it needs

Katy: MEX seems to be the right place to do this work

Anish: Given that this is a proprietary spec, I don't know what version is considered
... also how does it deal with attaching a policy to an epr?
... does mex deal with packaging a policy with an epr?

katy: Yes it deals with scoping

<David_Illsley> MEX 1.1 - http://specs.xmlsoap.org/ws/2004/09/mex/WS-MetadataExchange.pdf - Section 6 Metadata in Endpoint References

Anish: What is the timeline?

Katy: I will try to find out.

Ram: take a look at mex 1.1 and see if it takes care of the problem

<TRutt_> +1 with anish comment, Mex is not yet available

<David_Illsley> > MEX 1.1 - http://specs.xmlsoap.org/ws/2004/09/mex/WS-MetadataExchange.pdf - Section 6 Metadata in Endpoint References

Anish: It is not clear if and when it might be available to us.

Ram: two questions, 1) is it the right way to go and 2) what it the timeline

Anish: I don't think that the two questions are independent.
... Past history does not make me confident that it will be timely

Ram: is this a problem that this wg should do or not?

Anish: At lease some(many?) that think it should be done here

Gil: Given the history of WS-Policy, I understand Anish's concern

resolution: defer decision until after cr

<TRutt_> WS-policy Task Force Analysis: http://lists.w3.org/Archives/Public/public-ws-policy/2007Apr/0022.html

<TRutt_> From ws-policy framework 3.2 “ Definition: A policy alternative vocabulary is the set of all policy assertion types within the policy alternative.] When an assertion whose type is part of the policy's vocabulary is not included in a policy alternative, the policy alternative without the assertion type indicates that the assertion will not be applied in the context of the attached policy subject.”

TomR: above point is in discussion within WS-Policy
... Alternative H is a brute-force method that skirts the negation issue
... I think that we need to wait until WS-Policy decides

Anisk: Consider negation and the none uri as separate issues
... do you think that your response works for both types.

TomR: Depending on how the negation issue will temper which resolution we pick

Anish: Is it what constitutes a vocabulary in general or is it related only to nested assertions?

TomR: Most of the problem is from nested assertions
... there are also issues with regard to the definition of vocabulary
... this may be a vocabulary scoping rules

Anish: Some policy wonks say that there is no negation, just something is not defined

Policy subject (viz Anish's email)

<anish> here is the email with my 2 issues: http://lists.w3.org/Archives/Public/public-ws-addressing/2007Apr/0010.html

<TRutt_> There was a question from ws-policy members on the conformance to ws addressing, with respect to support for types of responses

<dhull> +1 (at least)

Anish: we do not define what wsam means
... what does it mean to assert the wsam: assertion
... Second, then is a statement that such an assertion may not apply to a port type

<gpilz> +1

Anish: Does it mean soap binding, does it mean core?
... What if I want to not support the none uri (one way messafe with faults, for example)

Ram: Alt G is a solution that defines what we need

<TRutt_> The "none" uri is just a special way to say that a partcular message is purely one way, it can be used regardless of what "response" types are supported/required for messages which expect a response of fault. I think we can word things to get around this none problem

Anish: Even in alt G does not specify if the soap binding is used or not. i.e. what spec are we making an assertion concerning

TomR: It is important to know what the client can do. Anish, please clarify what you mean by none

Anish: what assertion are you making and what spec does it apply to

Ram: we can consider both possible outcomes from the WS-Policy froup
... if negation does not exist, would that take away our concern about none?

<TRutt_> The 'None" uri implies that no response is expected. We could define the other policy assertions to only apply for cases where responses are expected, It is a matter of how we word the assertions

Anish: For me it would

Ram: What is the right thing in our opinion?

David: I assumed that the assertion applied to the subject, that means the core spec would be used and applied to the appropriate binding

Gil: I am -1 on separate abstract and non-abstract assertions

Anish: Are you proposing a context dependant assertion?

Gil: That is pretty much it
... I don't agree to the restriction prohibiting an abstract assertion
... I also don't like to define the assertion to apply to only one spec.
... spelling it out and tying it to specific documents is a good thing to do

<scribe> ACTION: gil and anish to work up a joint proposal [recorded in http://www.w3.org/2007/04/16-ws-addr-minutes.html#action01]

<gpilz> I thought so

<gpilz> then we should

<gpilz> my main point is that WS-Addressing is supposed to be a general purpose facility

<gpilz> that being the case, I can't see why we should define per-binding assertions to indicate a requirement that it be used

<gpilz> obviously, if you are using SOAP

<gpilz> and you say you are using addressing

<dhull> I'm just saying that we use "engaged" to describe a similar concept in one of our specs.

<gpilz> I'm not sure that anybody who isn't a standards person would have a hard time understanding this

<David_Illsley> hehe

<anish> :-)

<gpilz> here's an analogy: YOU MUST OBEY ALL TRAFFIC LAWS

<gpilz> does that mean I should obey the laws that apply to cars, bikes, or walking?

<gpilz> obviously, if I am driving a car, I should obey those laws that apply to cars

<gpilz> right?

Next meeting will be April 23, but if the WS-Policy has not clarified its position on negation, then it might be canceled if no other issues arise.

Summary of Action Items

[NEW] ACTION: gil and anish to work up a joint proposal [recorded in http://www.w3.org/2007/04/16-ws-addr-minutes.html#action01]
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.128 (CVS log)
$Date: 2007/04/24 00:03:42 $