XMLP WG Telcon Minutes 14 April 2004

Based on IRC log

Present 9/7
BEA Systems, Mark Nottingham
Canon, Herve Ruellan (scribe)
IBM, John Ibbotson
IBM, Noah Mendelsohn
IBM, David Fallside
Nokia, Michael Mahan
Oracle, Anish Karmarkar
SeeBeyond, Pete Wenzel
W3C, Yves Lafon

BEA Systems, David Orchard
Canon, Jean-Jacques Moreau
Oracle, Jeff Mischkinsky

Microsoft Corporation, Martin Gudgin
Microsoft Corporation, Jeff Schlimmer
SAP AG, Gerd Hoelzing
Sun Microsystems, Tony Graham
Sun Microsystems, Marc Hadley
Systinet (IDOOX), Jacek Kopecky

IONA Technologies, Suresh Kodichath
SAP AG, Volker Wiechers
Systinet (IDOOX), Miroslav Simek

[herve_scribe] * Approval of minutes of March 31
[herve_scribe] No corrections
[herve_scribe] No objection for approving the minutes

[herve_scribe] * Action items review
[herve_scribe] As in agenda, except review of WS-Reliability that is done.
[herve_scribe] Question about Mark's extensibility text.
[herve_scribe] Mark: waiting for up to date documents
[herve_scribe] Hervé: will edit MTOM beginning of next week

[herve_scribe] * Status reports

[herve_scribe] - Media types registrations
[herve_scribe] Mark: Still waiting for XML 1.1 decisions

[herve_scribe] - FAQ for XInclude
[herve_scribe] DF: No feedback from core wg
[herve_scribe] MM: Prefer current format (i.e. no collapsing)
[herve_scribe] MM: Ok for dropping Q9.
[herve_scribe] MM: concurs with 3/4 of what Jacek said.
[herve_scribe] MM: Ok for dropping reason in Q6 (was there mainly for completeness).
[herve_scribe] Note: MM intervention was about responding to Jacek comments.
[herve_scribe] DF: suggest to drop Q9.
[herve_scribe] DF: the FAQ should be a WG note. Therefore generate something like a ed copy of a WG note.
[herve_scribe] Yves: it could simply be a plaintext FAQ on W3C site.
[herve_scribe] DF: put it in the form of a WG document.
[herve_scribe] MM: will do it for next telcon.

[herve_scribe] - XMLP/WSD TF
[herve_scribe] Anish: did not manage to organize telcon.
[herve_scribe] Anish: 3 possibilities on solving the problem. TF will have to continue discussion.
[herve_scribe] DF: is there a schedule for telcon?
[herve_scribe] Anish: I will try to organize it, otherwise continue discussion through mailing list.
[herve_scribe] DF: is there a document now instead of the mail message version of the 'document'?
[herve_scribe] Anish: Yes.
[herve_scribe] Anish: I will find most recent version of the doc. Not sure the most recent version has been sent on mailing list.

[herve_scribe] - WS-Reliability spec
[herve_scribe] DF: Postpone this until PeteW join the telcon (see below).

[herve_scribe] * Attachments

[herve_scribe] - Representation Header doc
DF: regarding comments on the current draft ....
[herve_scribe] DF: suggestion from Anish: reinsert with no namespace. Do this change before publishing (first) WD.
[herve_scribe] DF: any other comment or thing to do before publication?
[herve_scribe] No answer.
[herve_scribe] DF: there was an action to the WG for solving 455.
[herve_scribe] DF: I tried to combine all pieces of text for generating resolution.
[herve_scribe] DF: I think we agreed (last telcon) to use Noah's text?
[herve_scribe] Noah: my proposal was nearly editorial.
[herve_scribe] DF: so we agree that my email contains all resolution pieces?
[herve_scribe] Noah: I think we need to say something about the atypical case?
[herve_scribe] Noah: i.e.: two headers with same metadata, what happens?
[herve_scribe] Noah: solutions are undefined, implementation dependant, first one wins...
[herve_scribe] Noah: I think we should say something.
[herve_scribe] Anish: should we allow people to do this?
[herve_scribe] Noah: we can forbid it, but should define the process with any kind of different input that it may receive.
[herve_scribe] DF: proposal representation header block SHOULD NOT same metadata. If the have the same metadata, then behaviour is undefined.
[herve_scribe] Noah: suggestion: either header block can be used.
[herve_scribe] DF: friendly amendment accepted.
[herve_scribe] DF: any more opinions?
No answer
[herve_scribe] DF: any objections?
[herve_scribe] No objection.
DF: we will use the revised proposal
[herve_scribe] Anish: shall I include this text in the upcoming WD?
[herve_scribe] DF: yes. Not much changes to do.
[herve_scribe] DF: that completes all we need to do before sending doc to W3C for publication as a WD.
[herve_scribe] Anish: I can update the document today if somebody can send text.
[herve_scribe] DF: I will send text.

[herve_scribe] - rec20
[herve_scribe] DF: we agreed at Infoset any version of XML, but nodes may be more restrictive.
[herve_scribe] DF: at binding level, media type is application/xml-soap, but other media types may be less restrictive.
[herve_scribe] Noah: are you saying that the media-type description contains the restriction and would apply to any binding using this media-type?
[herve_scribe] DF: correct.
[herve_scribe] DF: was are the subparts of the issue to deal with?
[herve_scribe] Noah: do we need new errors or faults?
[herve_scribe] Noah: for HTTP binding, for old style node would need new media type and node would not support it; error is defined.
[herve_scribe] Noah: if application has DOM parser or something that do not support XML 1.1?
[herve_scribe] Noah: if I write new binding was kind of fault is generated is there is a mismatch?
[herve_scribe] Noah: not sure what we have is ok.
[herve_scribe] DF: discussion?
[herve_scribe] Yves: need to tackle the issue in binding framework to ensure that right information is given in binding definition.
[herve_scribe] DF: need someone to check if binding framework is ok.
[herve_scribe] Yves: will do this.
[herve_scribe] Noah: and do we need some liaison with WSDL?
[herve_scribe] Noah: not possible to describe XML 1.1 using XML Schema, so new messages are not describable with WSDL.
[herve_scribe] Yves: WSD bound to support SOAP 1.2. And SOAP 1.2 support everything in Infoset, so this is nothing new.
[herve_scribe] Noah: This was buggy. On the legalistic point of view, it was clear. But informally not so.
[herve_scribe] DF: seems worthwile to tell WSDL about changes (accepting any version of XML).
[herve_scribe] DF: it may be possible to close rec20 next week (with Yves proposal).
[herve_scribe] DF: is proposal to add clarification text?
[herve_scribe] Yves: think so.
[herve_scribe] DF: might put something in HTTP section and application/xml+soap.
[herve_scribe] Noah: first thing is what Infoset enveloppe is.
[herve_scribe] Noah: our Infoset MUST be something obtained in parsing some version of XML.
[herve_scribe] DF: start generating those pieces of text?
[Yves] http://lists.w3.org/Archives/Public/xmlp-comments/2004Mar/0012.html
[Yves] (the middle link) of rec20
[herve_scribe] Yves: rec20 is 1.0 against 1.1, while rec22 is about all characters allowed in Infoset.
[herve_scribe] Yves: I think we tackled rec22 and that resolution contains resolution for rec20. rec20 is a specialized case of rec22.
[herve_scribe] Mark: we will need to discuss the impact of this on the media-type registration.
[herve_scribe] Yves: will be part of resolution.
[herve_scribe] DF: can talk about it now.
[herve_scribe] DF: registration of application/xml+soap
[herve_scribe] Mark: the registration can refer to or contain text in registration proposal.
[herve_scribe] Noah: points to RFC 3023 which allow any version, but can be further restricted by other specs such as registration.
[herve_scribe] Mark: I have enough information to generate a proposal.
[herve_scribe] DF: and what about the XOP media-type?
[herve_scribe] Mark: what do we want to call it?
[herve_scribe] Noah: call out presence of SOAP envelope in root?
[herve_scribe] Mark: recommend application/soap_xop+xml
No other names proposed, no objection to using that name.
[herve_scribe] DF: Mark, go ahead and write a proposal using that name.

[herve_scribe] - Rec22
[herve_scribe] DF: what more to close Rec22?
[herve_scribe] Yves: with all the text we will have, should be ok.
[herve_scribe] Yves: would need to double check, but pretty sure.

[herve_scribe] - 458
[herve_scribe] DF: we need to check whether or not 458 is a duplicate issue to Rec20/22.
[herve_scribe] Yves: its context is XOP/MTOM, therefore slightly different.
[herve_scribe] Noah: Don't see anything not covered.
[herve_scribe] DF: we did not talk about text changes for XOP/MTOM.
[herve_scribe] Noah: problem is XOP is independant of SOAP.
[herve_scribe] Noah: what shall we say about XML 1.1 in XOP? Might be looser than in SOAP.
[herve_scribe] Noah: XOP should allow any version of XML, soap_xop media type will restrict things.
[herve_scribe] DF: include this in XOP/MTOM drafts.
[herve_scribe] DF: same thing to do with representation header, but after publication (for not delaying it).

[herve_scribe] * Status reports
[herve_scribe] - WS-Reliability
[herve_scribe] Pete: List of comments and issues.
[herve_scribe] Pete: 1- Lack of handling of intermediaries in the spec.
[davidF] pete's comments at http://lists.w3.org/Archives/Member/w3c-xml-protocol-wg/2004Apr/0019.html
[herve_scribe] Pete: Do not target intermediaries, but nothing prevents targeting reliability header to them.
[herve_scribe] Pete: 2- Provide extensions for 1.1 and 1.2, 1.1 non-normative, while 1.2 normative, but all examples use 1.2.
[herve_scribe] DF: we should push to use 1.2.
[herve_scribe] DF: does this represent group opinion?
[herve_scribe] Yes.
[herve_scribe] Pete: use term Reliable Messaging Processor, and use sending/receiving RMP.
[herve_scribe] Pete: Do we have another term for this?
[herve_scribe] DF: could ask them to be more consistent about terminology.
[herve_scribe] Pete: RMP includes SOAP Processor.
[herve_scribe] DF: ask them if then can be clear about use of terms module, node, processor.
[herve_scribe] Pete: SOAP Fault can be used to convey a reliable messaging acknowledgement.
[herve_scribe] Pete: is this ok?
[herve_scribe] DF: put opinion into a note.
[herve_scribe] DF: send draft by eod tommorrow and if not comments, send it on Monday.
[herve_scribe] DF: if there are comments, will have to decide whether to send it or not.

[Zakim] WS_XMLP()11:30AM has ended

[RRSAgent] I see 12 open action items:
[RRSAgent] ACTION: Michael to "update" the FAQ and turn it into a WD in time for April 21 telcon [1]
[RRSAgent] ACTION: chair to contact Core re publishing xbinc:include FAQ [2]
[RRSAgent] ACTION: anish to send chair a ptr to most uptodate version of media type doc [3]
[RRSAgent] ACTION: DavidF to send final resolution text for issue 455 [4]
[RRSAgent] ACTION: yves to evaluate binding framework for sensitivity to XML version, in time for April 21 telcon [5]
[RRSAgent] ACTION: yves to send heads up to WSD WG re. our interpretation of infoset description in SOAP 1.2 [6]
[RRSAgent] ACTION: noah to draft SOAP 1.2 errata text based on our provisional resolution to rec20, by end of week [7]
[RRSAgent] ACTION: markN to generate registration proposal for app/soap+xml incorp our proposed resolution to rec20, due monday april 19 [8]
[RRSAgent] ACTION: markN to generate registration proposal for app/soap_xop+xml, due monday april 19 [9]
[RRSAgent] ACTION: yves to double check that resolution materials being created fro rec20 will also cover rec22 [10]
[RRSAgent] ACTION: herve looak at noah's SOAP 1.2 errata text and propose changes for MTOM [11]
[RRSAgent] ACTION: peteW to draft response to WSRM TC by CoB tomorrow, and to send response on monday to TC unless critical f/back received [12]