XMLP minutes

14 September 2005


See also: IRC log

1. Roll Call

BEA Systems, David Orchard
Canon, Herve Ruellan
IBM, Chris Ferris
Nokia, Mike Mahan
Oracle, Anish Karmarkar
SeeBeyond, Pete Wenzel
Sun Microsystems, Marc Hadley
W3C, Yves Lafon
IBM, Noah Mendelsohn
BEA Systems, Mark Nottingham
Iona Technologies, Suresh Kodichath
Microsoft Corporation, Mike Vernal
SAP AG, Volker Wiechers
Canon, Jean-Jacques Moreau
Microsoft Corporation, Doug Purdy
Oracle, Jeff Mischkinsky
Sun Microsystems, Tony Graham
Mike Mahan

2. Agenda Review , Announcements, and call for AOB

Will discuss logistics at end of meeting.

3. Approval of Minutes

17 Aug: http://lists.w3.org/Archives/Member/w3c-xml-protocol-wg/2005Sep/att-0002/2005-08-17-minutes.html

4. Action items

Action Items: http://www.w3.org/2000/xp/Group/Admin/#pending

2005/08/10 Yves
Send a proposal to resolve issue rec33
Yves: no time to do it. Maybe next week.
Marc may have sent something, need to look for it.
ACTION: Marc to find rec33 related email thread [recorded in http://www.w3.org/2005/09/14-xmlprotocol-minutes.html#action01]
Mike: OK lets take this on the ML and discuss next week.
Disposition: still PENDING

2005/08/17 WG
Evaluate the possibility of reviewing appendix K and/or L from Voice Browser
Mike: any volunteers to do this review?
Mike: OK – I will do it.
ACTION: Mike to review voice browser appendices [recorded in http://www.w3.org/2005/09/14-xmlprotocol-minutes.html#action02]
Disposition: close original ACTION, new ACTION to Mike

2005/08/17 WG
Send out a revised timeline for the new XMLP charter draft
Mike: still noodling on it. AI remain pending. We will address this later in this meeting
Disposition: still PENDING

2005/08/17 Chris
ATF Issue: Does the SOAP/HTTP binding require a SOAP env in the response?
Continue discussion, and send concerns about 202 and different MEP levels in the ML
Mike: close this AI with Chris and Pick it up with Marc's AI.
Postlude: Above statement is confusing issue rec33 with this ATF issue
Disposition: Keep PENDING, Mike to check progress of this issue

2005/08/17 Editors
Summarize the current requirements for the one way MEP
Disposition: Keep PENDING, Editors are tasked to come to an approach concensus. These are closely related.

5. XMLP Requests

1. Voice Browser WG
See: http://lists.w3.org/Archives/Member/w3c-xml-protocol-wg/2005Aug/0019.htm
mikem: Voice Browser WG wants XMLP to review bindings of their infoset to HTTP
mikem: anyone familiar ? volunteers ?
mikem: Yves, is it mandatory, or optional, to review ?
yves: we can decline if we decide that no useful comment can be made
yves: one comment might be to recommend that they use SOAP
mikem: is there a timeframe?
yves: Voice Browser group is on holidays during august
mikem: action to WG to consider reviewing either appendix K or L or both, we will bring it up in 3 weeks time

2. ATF Issue: Does the SOAP/HTTP binding require a SOAP env in the response?
Proposed response: http://lists.w3.org/Archives/Member/w3c-xml-protocol-wg/2005Aug/0015.htm
anish: explains the history on this. MarcH's has a proposal to say what needs to be changed to send an empty reponse
chris: notes that MarcH's proposal would not accomodate the likes of use of HTTP response for ReliableMessaging ack.
marcH: concurs
anish: notes that he had similar concerns were raised before
anish(IRC): here is what HTTP spec says about 202: "The entity returned with this response SHOULD include an indication of the request's current status and either a pointer to a status monitor or some estimate of when the user can expect the request to be fulfilled."
noah: agrees with ChrisF and Anish. MEP purpose is to say exactly what is expected.
chris: discusses whether application level MEP or the underlying mechanics (plumbing) should handle this
anish: one of the problems that exists, as i see it is, there are too many patterns: transport-level MEPs, SOAP MEPs, WSDL MEPs, Application MEPs. They are all layered over each other. With optionality introduced, there are all kinds of interactions between them and causes a problem with mappings.
noah: SOAP processing MEP should say how to process 202
davido: SOAP spec. has 2 different state machines for 2 MEPs, while http binding has one state to deal with it.
noah: the real test for the MEPs is in using it with SOAP intermediates
ACTION: Chris to continue discussion, and send concerns about 202 and different MEP evels in the ML

6. New SOAP MEP/Binding work item

1) Charter has feedback from W3C, WSD.
Typo found by Noah
Hugo: http://lists.w3.org/Archives/Member/w3c-xml-protocol-wg/2005Aug/0024.htm
from WSD: http://lists.w3.org/Archives/Member/w3c-xml-protocol-wg/2005Aug/0025.htm
from Yves: http://lists.w3.org/Archives/Member/w3c-xml-protocol-wg/2005Aug/0026.htm
Mike: Some comments on the charter have been sent.
... from Hugo, OK, but like to see explicit requirements like the language from WS-D.
... from WSD: more explicit deliverables preferred
Yves: the deliverables are clear for one-way
Mike: What does the rest of the W3C team think
Yves: Phillipe thinks it is good
Yves: Need to have the new charter fast, as current one is about to expire.
... Timeline is very aggressive. May need to push it back 3-6 months. Depends on available time from WG.
Mike: other comments on timeline?
... Timeline correct or need modifications?
Herve: Pushing back a month makes fo the WD and LC
No response.
<Yves> http://lists.w3.org/Archives/Member/w3c-xml-protocol-wg/2005Aug/att-0011/XML_Protocol_WG_Charter.htm
Yves: question is about first WD in November and LC in December. Rest will follow.
Mike: what about reqs from WSA? Will they only be additional or could they be contraditory to what we already have?
Marc?: WSA reqs will not be contradictory with WSD's ones, but may broaden some scope.
Anish: If we start now, with what to start?
... can start talking about details, event if we don't know exactly were we go.
Concern: dependencies with WSA and WSD.
Yves: may delay 1 step by 1 month.
... delay CR by 2 months.
... will not be a problem as WSA and WSD will be able to refer to the spec.
Mike: Yves, so you believe a 3 months delay is OK for WSD?
Yves: think that this is a reasonable delay.
Mike: what about dependencies with WSD and reqs from WSA.
Yves: should check with Hugo, but think it will be ok.
<Yves> ACTION: Yves to check with Hugo wrt timeline and dependencies with WSA and WSD [recorded in http://www.w3.org/2005/09/14-xmlprotocol-minutes.html#action03]
Mike: anything else about charter and feedback.

2) Scoping of new work item
Lets take up where we left off on Aug17
See min: http://lists.w3.org/Archives/Member/w3c-xml-protocol-wg/2005Sep/att-0002/2005-08-17-minutes.html

Mike: editors to put a new proposal, in order to be proactive and help WSA to make their requirements.
Chris: already have some proposal from Dave. Is this sufficient.
... not clear to me if WG has consensus on the approach.
Mike: proposal is for the editor to come up with a sketch and give it to the WG to see what we have.
... also give it to WSA to help them make their reqs.
Chris: have been discussing several approaches.
Anish: WG hasn't come to consensus about how thing should work. Problem is when as message is going out, but nothing coming back.
... WG should look at this to achieve a consensus.
... we know what we want, we have at least one proposal. We can look at the details. We can agree on how it will work, the spec may come later.
Mike: Proposal: describe in the WG all the details without working on the spec.
Anish: should not wait too long for WSA.
Mike: Scoping of what we will do is limited by reqs from WSD.
Chris: In line with what I was suggesting.
Mike: working on that and expanding it if we get sufficient feedback from WSA.
<dorchard> http://www.pacificspirit.com/Authoring/async/async-scenarios.html#twoway-inonly
Dave: no consensus on how many MEP. However there are some details that will be present whatever the approach.
<cferris> personally, I think NO MEPs :-) Just a binding
Dave: would like to propose a sketch of what we will do and ask for feedback.
... need to have an agreement between the editors.
Chris: Fundamentaly we have SOAP MEP, WSD MEP, WSA MEP. Ultimately there should be no SOAP MEP.
<marc> amen brother
Chris: MEP are at the application level. Preference would be to do only a binding: here is how SOAP is transported on HTTP.
... Then can compose with WSD MEP.
Dave: This is the approach I propose. The only reason to have a notion of MEP is to have a URI to refer to in the request.
Chris: Think there is no consensus on it.
<marc> dave characterised this as request, optional response. i think optional request, optional response is closer
<marc> this will support the existing GET behaviour
<dorchard> think that we need to have consensus amongst editors first, then put to wg.
Dave: If editors have a consensus, then may propose something to WG.
Mike: Editors should be chartered with crafting the approach and presenting it to WG.
<dorchard> So let's get Chris, myself and Anish together and bash this out.
Anish: if consensus happen by next telcon, would be great. Need to go deeper in the details. Need to have a short concall.
Mike: Editors to come to consensus for next week.
Mike: Editors meet at their discretion for this
Editors: Arrange da phone conference

7. SOAP 1.2 PER specs [Did not address]

1. 33rec http://www.w3.org/2000/xp/Group/xmlp-rec-issues.html#x33
2. Behavior of Requesting node on a 3xx response in the SOAP HTTP inding
See: http://lists.w3.org/Archives/Public/xmlp-comments/2005Aug/0000.html
& http://lists.w3.org/Archives/Public/xmlp-comments/2005Aug/0003.html
3. Does the SOAP/HTTP binding require a SOAP env in the response?
Anish: http://lists.w3.org/Archives/Public/xml-dist-app/2005Aug/0011.html
Anish: http://lists.w3.org/Archives/Public/xml-dist-app/2005Aug/0012.html
Marc: http://lists.w3.org/Archives/Public/xml-dist-app/2005Aug/0014.html

8. XOP Issues - Pending, not exhaustive [Did not address]

1.34rec start-info: http://lists.w3.org/Archives/Public/xmlp-comments/2005Aug/0001.html
2.35rec Use absolute URIs for action: http://lists.w3.org/Archives/Public/xmlp-comments/2005Aug/0002.html
3. NEW Issue: Possible defect: http://lists.w3.org/Archives/Public/xmlp-comments/2005Aug/0004.html

9. AOB

Will be moving time back to old start time – 30 minutes earlier
- Moved up without comment from European members
- We are still using a minimum of 1.5 hrs, 1 hr doesn't seem to be sufficient

[NEW] ACTION: Marc to find rec33 related email thread [recorded in http://www.w3.org/2005/09/14-xmlprotocol-minutes.html#action01]
[NEW] ACTION: Mike to review voice browser appendices [recorded in http://www.w3.org/2005/09/14-xmlprotocol-minutes.html#action02]
[NEW] ACTION: Yves to check with Hugo wrt timeline and dependencies with WSA and WSD [recorded in http://www.w3.org/2005/09/14-xmlprotocol-minutes.html#action03]