Web Services Addressing WG Teleconference

21 Feb 2005



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




Agenda review, AOB

Jonathan: I was wondering which issues we could leave open as we take Core and SOAP binding to LC
... I'd like to walk into the F2F knowing which issues we need to close

Mark: we'll be discussing this later in this call then

Call for corrections to the minutes

Minutes are at http://www.w3.org/2002/ws/addr/5/02/14-ws-addr-minutes.html

Mark: Umit wanted to clarify something she said

Minutes accepted with Umit's amendment

Review action items

Mark: I saw that Gudge did the schema AI
... did you see the original version of the schema?

Gudge: sorry, I didn't and started from scratch
... we will approve it next week

Jonathan is reminded of his IRI AI

Upcoming Meetings

Mark: we'll be meeting Sunday-Tuesday next week
... we will go to Bob's for dinner on Sunday

Bob goes over the yummy menu (including stinky cheese)

Bob, Mark and Hugo discuss details

An email will be sent out to the WG

Mark: you will need somebody to get in, so please be prompt
... call Philippe or myself if you're late
... we'll have a meeting with the TAG; Paul will be presenting Addressing and our issues to the TAG

Paul: I'll try to have slides early
... I'd be interested in pointers from people

Mark: with regards the April F2F, we're looking at 19-20 April
... having not heard any push back, those dates are now final
... we'll be sending admin details soon
... for June, we're still looking at Berlin on 30 May
... we're looking at co-hosting with WSDWG again

Jeff: I'd prefer the end of the week

Proposed and New Issues

Hugo presents Proposed: Misalignment of treatment of reply messages and fault messages; http://www.w3.org/mid/20050215151434.GD13607@w3.org

Issue accepted; owner is Hugo

Friendly amendment is: http://lists.w3.org/Archives/Public/public-ws-addressing/2005Feb/0102.html

Mark: please have a look at this this week, and we can close this next week

<mnot> new modules: http://www.w3.org/mid/20050218154320.GE10971@w3.org

Hugo presents Definition of SOAP 1.2 (and 1.1) modules; http://lists.w3.org/Archives/Public/public-ws-addressing/2005Feb/0118.html

Mark: I'd like to make this a proposal for issue i022

Hugo presents "Binding of message addressing properties in the SOAP underlying protocol"; http://www.w3.org/mid/20050218154335.GF10971@w3.org

Mark: do you want this to be folded under i022?

Hugo: no preference

Greg: I prefer a separate issue

Hugo is the owner of the new issue

<mnot> proposed issue: what is a logical address? http://www.w3.org/mid/421A36B2.6050309@oracle.com

Anish presents "What is a logical address?"; http://lists.w3.org/Archives/Public/public-ws-addressing/2005Feb/0132.html

Mark: any problem with putting this on the issues list?

Gudge: what's the difference between logical and dereferenceable?

Jeff: what's the difference between logical and physical?

Gudge: that the address may not be the physical one

<TomRutt> I agree this is an important issue as the spec stands

<pauld> for the most part liked anish's proposal, though i'm unsure about what dereferencable really means.

Anish: I'm getting a lot of questions about "logical address"
... we should at the minimum define it

Dave: WebArch defines what dereferenceable means, so I think it would be good to reuse the term, and I do agree that there is some confusion around this

<uyalcina> you missed me David. I know what deferenceable means. I am just trying to understand the logical address vs. a dereferanceable address

Umit: I agree there is an issue

Tom agrees

Issue is added to the issues list; Anish is the owner

<uyalcina> there is an issue with the wording in the text with respect to what logical address means.

<pauld> just looked at the web-arch, using 'dereferencable' looks good!

Anish: I will add more references, e.g. to WebArch in my proposal

What issues do we need to close in order to go to LC for Core & SOAP binding?

Mark: my thinking was that we needed to close issue i004, i007
... i017 is WSDL specific
... what about i020?

Anish: I think that this is WSDL specific

Mark: i021 is WSDL specific
... i022 need to be closed
... so do i024 and i26
... i041 can wait
... i042 and i43 need to be closed
... i048 too
... all the new issues need to be closed

Jonathan: what can we do about i022?

Mark: it's unclear right now

Greg: i022 certainly has a relationship with the WSDL document

Mark: on the last call, the TF thought that we didn't need any big changes

i004 - Security Model

Mark: we're missing an AI from Marc

<mnot> http://www.w3.org/mid/DD35CC66F54D8248B6E04232892B633804D931FF@RED-MSG-43.redmond.corp.microsoft.com

Gudge presents http://lists.w3.org/Archives/Public/public-ws-addressing/2005Feb/0130.html

Mark: is there a relationship with i007 with regards to the text about intermediaries?

Gudge: my understanding is that i007 is about targeting nodes, not tampering with default values and XML representations

Anish: this isn't specific to Addressing, right?

Gudge: the particularity is default values

Gudge: these are our headers

Hugo: if all specs defining headers need to do that, is there anything wrong with SOAP?

Gudge: we have default values, which is where the issue is
... to work around this, we would need to define a C14N transform

Anish: does that only apply to forwarding intermediaries?

Gudge: agreed

Mark: are people comfortable with this? could we close i004 with this?
... I know Marc and Paco had other ideas in mind

Considering Gudge's proposal with Anish's amendment about forwarding intermediaries

Greg: do we say that it means to sign an EPR?

Gudge: no, because I couldn't get some agreement around that with the people I talked to
... there are different feelings about what needs to be done to secure EPRs

Mark: one thing to consider is whether this is good for us; if we need more than this, we may be able to do something as an extension later
... let's wait for Marc and Paco to see if they want to add something to this; we'll decide on this early in the F2F

i042 - Extensibility Model

Jonathan presents http://www.w3.org/mid/7DA77BF2392448449D094BCEF67569A50670C4D6@RED-MSG-30.redmond.corp.microsoft.com

<mnot> http://lists.w3.org/Archives/Public/public-ws-addressing/2005Feb/0019.html

Mark: would anyone object to closing i042 with this proposal?


RESOLUTION: Issue i042 closed with http://lists.w3.org/Archives/Public/public-ws-addressing/2005Feb/0019.html

<scribe> ACTION: Editors to integrate http://lists.w3.org/Archives/Public/public-ws-addressing/2005Feb/0019.html

i049 - 'default default' action URI for fault messages

Marc hasn't done his AI

Chair will ping Marc

i007 - Processing model of headers

<mnot> http://www.w3.org/mid/20050218154308.GD10971@w3.org

Hugo made a proposal for addressing the first part of the issue: http://lists.w3.org/Archives/Public/public-ws-addressing/2005Feb/0117.html

Mark: do people think that targetting is a good idea?

Anish does

<mnot> http://www.w3.org/mid/80A43FC052CE3949A327527DCD5D6B27D1B5F6@MAIL01.bedford.progress.com

Mark: Glen has a proposal, but it's not detailed enough IMO
... would anyone would like to expand on this proposal?

Jonathan: I'm confused about it

<scribe> ACTION: Hugo to make his proposal about i007 clearer

i022 - Relationship to the SOAP binding framework

Mark: we have a TF looking at the relationship between SOAP, Addressing, and WSDL
... Hugo raised 2 related issues
... what do we need to do to close i022?

Greg: I want to know what portion goes as SOAP messages, and what portion goes as transport messages [scribe poorly paraphrasing]
... I am wondering if we should move a portion of the SOAP document in the WSDL document
... that may mean being able to delay the decision on i022 beyond next week

Mark: you would like to have transport-specific wording in the SOAP spec?

Greg: yes
... I have heard different things about the current SOAP binding
... I think that we need to call a specific behavior out

Anish: do you think that the binding satisfies what we need?

Greg: yes

Anish: my interpretation was that, for SOAP 1.2, the current transport binding wasn't enough

<uyalcina> +1 to Anish, this is my understanding as well

Anish: for 1.1, because of the WS-I interpretation, one-way is covered
... this isn't the case for 1.2

Greg: but this is for one-way, not for req-resp

Umit: the discussion was around whether the WSDWG should define a new SOAP binding

Greg: I'm hoping that we can discuss options at the F2F, and choose the bare minimum to get out of this

Mark: are you talking about LC or CR?

Greg: CR; it would be nice to have them fixed for LC [confusion around LC/CR]

Mark: do people think that we can ship documents as LC drafts with the resolution of these issues?
... if not, can we have this solved in other drafts?

Jonathan: I think we'll be able to ship our LC drafts
... I'm not sure that we would even need to publish other drafts as Recs

Umit: it seems to me that we have SOAP and WSDL issues, not Addressing ones

Mark: what do we need to do to close i022?
... Greg wants to see info from the TF
... Hugo wants us to define a SOAP module
... anything else?


Mark: have people had a chance to look at Hugo's proposals?

Jonathan: I don't see anything wrong with it, but I'd like a little more time

Mark: let's talk about it next week

i024 - Self-Contained vs. referencing EPRs

Mark: in the most recent proposal, I didn't see a way to differentiate referenced vs. included metadata

Rebecca: but you could put a URI in the metadata section

<TomRutt> the definer of the metadata namespace determines the semantics

Anish: I thought that the difference was service QName vs. service element

Mark: do we want to have a discussion about whether there should be a generic mechanism for doing self-Contained vs. referenced metadata?

Umit: it all depends on the element you put in there

Mark: Rebecca, would you be comfortable dropping i024 in favor of i026 which encompasses this?

<anish> i thought rebecca was interested in having the complete wsdl inlined in an EPR

Rebecca: no, we have a specific proposal

Mark: we would need a concrete proposal by next week then

Rebecca: would you like me to split the proposal i026 into 2?

Mark: if this is what you want, yes

Jonathan: what does it look like to reference metadata in your proposal?

Rebecca: we're proposing a generic metadata bucket

Umit: we're using QNames for service names, as an example

Mark: the way I understand i024, it's about providing a generic mechanism to refer to any type of metadata

<RebeccaB> Sorry - I just lost my phone connection

Anish: I remember that Rebecca wanted to have a whole WSDL in an EPR

Mark: my understanding was really about a generic mechanism

Jonathan: I'm confused; maybe we should be discussing i026 first

Umit: I think that Rebecca got it right; for i024, we're providing a generic metadata bucket; for i026, we're using it for multiple endpoints

Rebecca confirms

Mark: I think it would be helpful to have a separate proposal for i024
... otherwise it's hard to differentiate them

i026 - Supporting Multiple Ports in EPRs

Mark: we have a detailed proposal

Jonathan: I have ideas for a friendly amendment; I'd like to take it up next week

Dave: the notion about putting metadata in a metadata container doesn't make sense to me
... in order to differentiate data and metadata, we would need to have something in the spec to talk about how to use this
... we already have a pretty open extensibility model

Anish: we have wording in the spec (section 2.3) about metadata comparison
... we already have an issue on EPR comparison
... depending on its resolution, we would need this or not

Umit: I agree with Anish, and I would tend to agree with Dave too

Mark: maybe the authors of the proposal would want to provide an updated proposal

Anish: certainly, there is a dependency on the EPR comparison issue

I have made the request to generate http://www.w3.org/2005/02/21-ws-addr-minutes hugo

Summary of Action Items

[NEW] ACTION: Editors to integrate http://lists.w3.org/Archives/Public/public-ws-addressing/2005Feb/0019.html
[NEW] ACTION: Hugo to make his proposal about i007 clearer
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.111 (CVS log)
$Date: 2005/02/23 18:32:13 $