Web Services Addressing WG Teleconference

7 Feb 2005


See also: IRC log


Rebecca Bergersen (IONA Technologies, Inc.)
Francisco Curbera (IBM Corporation)
Glen Daniels (Sonic Software)
Paul Downey (BT)
Michael Eder (Nokia)
Robert Freund (Hitachi, Ltd.)
Yaron Goland (BEA Systems, Inc.)
Arun Gupta (Sun Microsystems, Inc.)
Marc Hadley (Sun Microsystems, Inc.)
David Hull (TIBCO Software, Inc.)
Yin-Leng Husband (HP)
Anish Karmarkar (Oracle Corporation)
Philippe Le Hégaret (W3C)
Mark Little (Arjuna Technologies Ltd.)
Jonathan Marsh (Microsoft Corporation)
Jeff Mischkinsky (Oracle Corporation)
Mark Peel (Novell, Inc.)
Tony Rogers (Computer Associates)
Tom Rutt (Fujitsu Limited)
Greg Truty (IBM Corporation)
Steve Vinoski (IONA Technologies, Inc.)
Pete Wenzel (SeeBeyond Technology Corporation)
Steve Winkler (SAP AG)
Ümit Yalçınalp (SAP AG)
Ugo Corda (SeeBeyond Technology Corporation)
Jacques Durand (Fujitsu Limited)
Martin Gudgin (Microsoft Corporation)
Eisaku Nishiyama (Hitachi, Ltd.)
Ales Novy (Systinet Inc.)
David Orchard (BEA Systems, Inc.)
Harris Reynolds (webMethods, Inc.)
Rich Salz (DataPower Technology, Inc.)
Davanum Srinivas (Computer Associates)
Jiri Tejkl (Systinet Inc.)
Hugo Haas
Mark Nottingham
Michael Eder



zakim unmute me

<scribe> scribe: MSEder

Agenda Item 3: Call for corrections to the minutes

Mark: no objection approving the minutes from January 24.
... approve the minutes from January 31?
... no objection

Agenda Item 4: Action item review

anish: did action item for number 17

Agenda Item 5: Co-ordination

mark: give TAG enough information to close their issue
... 90 minute joint meeting before lunch on Monday
... rough agenda, identity issues and specific issue that Mark brought up
... discussion about the issue that Mark Baker brought up

jonathan: faster we can make this go away the better. Has said everything he has to say on this issue

<GregT> ?

jonathan: not an issue all and cannot be solved in the way Mark Baker suggests
... growing what we think is the Web into new areas with what we call Web services
... does not have the time to give a presentation to the TAG

markn: should not dedicate too much time to this

Agenda Item 6: Future Meetings

markn: registration for the Sunday meeting and for Sunday dinner

Bob: there will be a bus provided

markn: bus cost will be about $20
... March in face-to-face agenda will be coming out soon
... no host for the April face-to-face
... April face-to-face should ideally be on the West Coast

<TomRutt> what are dates of april f2f

<Marsh> We might have to consider not co-hosting WSA and WSD meetings as well...

markn: only current offer is to host the meeting in Japan

<Marsh> ... if that increases the hosting possibilities.

Agenda Item 7: Working Draft Publication

markn: new drafts have been circulated

anish: what did we decide with respect to EPR comparison

markn: still not resolved

<tibco-dmh> +1

markn: what should we put in the status section of these documents

glen: work is ongoing this represents the current state of thinking, bla bla bla

markn: let the editors make the status section appropriate?
... no objection to publishing
... all three voted to be published as working drafts

Agenda Item 8: Proposed and New Issues

markn: new default default issue

jonathan: will take ownership of this item
... prepare concrete proposal

<scribe> ACTION: develop discussion and make a proposal for Jonathan

<anish> ACTION: develop discussion and make a proposal for Jonathan

Agenda Item 9: Issues

markn: no additional discussion on issue 4 security model
... people seem happy and we should say we have delivered a security modeling close issue out

paco: advisable to architect a way to sign the EPR's

markn: wants to avoid a comprehensive proposal right before the face-to-face
... going to close Gudges action item

<scribe> ACTION: Marc Hadley identify parts of the spect the place requirements are security model

markn: issue 17

anish: Anish and Jonathan have made proposals on this issue

<anish> http://lists.w3.org/Archives/Public/public-ws-addressing/2005Feb/0046.html

anish: two proposals are complementary

<anish> http://lists.w3.org/Archives/Public/public-ws-addressing/2005Jan/att-0010/00-part

anish: Jonathan talks about cross-referencing the two things in the two specs and Anish talks by using the best practice

jonathan: two proposals are if not complementary they are orthogonal

anish: operation name mapping best practice says
... if you have an action item that is distinct in a particular interface
... use WSDL extensibility

markn: is this only really valid at the interface level?

anish: could put this in other places as well
... this goes in our WSDL binding

jonathan: sense this is now a best practice what information does this give you?
... since this is not a requirement equals true can use the WSDL document

umit: not really sure what this proposal helps me do if I do not have these action values defined

anish: WSDL that you have will either have WSA action attribute if using addressing. You know what the action attribute is if you have addressing

jonathan: if not required this doesn't really help me just give me an opportunity to make errors

paco: if this problem has to be solved it will come from WSDL
... not our job to do what the WSDL working group should do

glen: WSDL says best practice
... best practice as if he do something that is hard to figure out that it is recommended you put an indicator in there to figure out what you're doing

anish: WSDL 2.0 had a requirement.
... this is a syntactic shortcut.

jonathan: two values for having the markup
... author made a mistake argument. Not very impressed with this one

markn: is this a well understood issue?

anish: if we have a WSDL 2.0 binding this should be done
... the best practice in WSDL 2.0 is to put in some marker
... would like to see and attribute in their.

jonathan: proposes an editorial solution
... wants cross-link between the two documents

markn: leave issue 17 open and direct the editors to include this text as an editorial text to the drafts.

<mnot> http://www.w3.org/mid/7DA77BF2392448449D094BCEF67569A50670C5A8@RED-MSG-30.redmond.corp.microsoft.com

<scribe> ACTION: Marc to incorporate the text in the above proposal

<Arun> Need to leave early so going offline

<scribe> ACTION: Jonathan coordinate regarding the WSDL half of his proposed text

<mnot> ACTION 3 = Marc to incorporate < http://www.w3.org/mid/7DA77BF2392448449D094BCEF67569A50670C5A8@RED-MSG-30.redmond.corp.microsoft.com> into editors' drafts

<mnot> ACTION 4 = Jonathan to coordination getting WSDL's half of < http://www.w3.org/mid/7DA77BF2392448449D094BCEF67569A50670C5A8@RED-MSG-30.redmond.corp.microsoft.com> into WSDL WG

markn: issue 18

<anish> http://lists.w3.org/Archives/Public/public-ws-addressing/2005Feb/0029.html

anish: only one place in the core were said properties are abstract
... not really sure if there is an issue

glen: using the reference properties to do some kind of extensibility thing

anish: another use case is a relay attribute

markn: options to leave it as it is, come up with some sort of abstraction with another attribute

ansih: there is no place were mislead people. neutral if we should add text

glen: no suggested text for 2.1, put it out and see if people get confused or not

markn: this seems to be a minor issue as some of the other points regarding endpoint should be captured as issue 20
... what is the resolution of this issue?

glen: rather say nothing at all then do something weird

<mnot> proposed resolution of i018: closed with no action; WG notes that multiple protocols can be accommodated by multiple attributes on RefPs, etc.

glen: no objections to closing with the above resolution

markn: issue 20

anish: talk about endpoints in WS addressing and have a definition of what endpoints mean
... WSDL 2.0 also has a definition of endpoint
... when we talk about endpoints and these two different specs they are quite different things
... proposal is to add a wording that says they're two different things

jonathan: Have three things, Web services endpoints, WSDL endpoint components, and we have endpoint references
... three different things and no overlap

ansih: the problem is different things but both talk about endpoints
... very different things. One should not confuse the two

jonathan: make the point that they are optimized for two different uses
... using the term endpoint too much
... should change the terminology in WSDL

glen: is an endpoint reference also an input description?

jonathan: agree the information describes the same behavior of the endpoint
... we optimize them for little bit different usage

paco: the EPR is runtime.
... endpoint component is more static
... one is a runtime artifact the other is a development artifact. not a black-and-white distinction

anish: discussion has gone on a tangent
... raise the issue that the terminology was confusing
... happy to give editors a license to correct this text

jonathan: want to see it before he signs it

anish: not much support to change for a name and would would rather see clarifying text

jonathan: make sure that WSDL talks about endpoint components

<scribe> ACTION: Anish will propose a text to clarify this issue

<anish> ACTION: Anish will propose a text to clarify this issue by 2/10

<anish> ACTION 5=

markn: issue 22

more a WSDL issue then addressing issue

scribe: anything in issue 22 that stops us from having last call on the SOAP binding

paco: do not know how we express this in WSDL
... and do not know how to bind SOAP

marc: is core ready

paco: For WSDL there's still work to do
... SOAP may be less of an issue
... allow core to finish

markn: needs a definite issues regarding soap and WSDL

greg: still need to understand better how we use these on the wire

paco: arbitrary how we're splitting the job of the group

anish: taking SOAP to last call at the same time as core

markn: do we need to talk about a particular transport binding in our document

ansih: there is not a soap 1.2 binding as a standard that can be used

markn: does not see how that stops us
... will have a chance in CR to see the specs integrated

greg: can we go to CR and still make changes

markn: if we need to make substantial technical changes we have to drop it back to last call

<pauld> observes this spec is based upon a member submission with existing implementations

markn: cannot exit CR if we do not show interoperability
... agree the use cases

paco: a big issue is on the table that we need to address

greg: we are working the issue through the task force

jonathan: not clear the solution space for the issue is going to touch the addressing core specification

markn: this issue will not affect the text of the core specification

paco: yes we agree that

marc: can we push the core ahead on its own and get it out of CR

markn: moved to the issues raised an issue 22 to the WSDL working group and they are a different schedule from us

<marc> not sure how we can exit CR with core alone, how do you test interop on an abstract thing

paco: dates are important but we need to make sure we do the work

markn: one of the reasons for last call is to have other working groups review the spec
... this will help to push other people to get their work done

ying-leng: what is the process for considering a something is major or not

philippe: no formal process
... but if it can change the opinions of reviewers of the document it can be considered important enough

markn: needed more discussion about the process for last call
... people mentioned things that need to be added to the soap binding draft


<anish> ACTION: glen to write up specific text to allign WS-Address to SOAP extensibility model

markn: issue 42

jonathan: put together a miniumal proposal for extensibility
... do we need something like a WSA must understand
... put in a boilerplates
... must ignore rule as opposed to a must understand
... extensions can introduce new properties

paco: explain point 3

jonathan: meant message properties

Minutes formatted by David Booth's scribe.perl version 1.109 (CVS log)
$Date: 2005/02/11 03:43:24 $