W3C XML Protocol Working Group teleconference, 21 April 2004

1. Roll

Present 12/10 Excused Regrets Absent

2. Agenda Review - now with 'important stuff'

 SOAP 1.2 errata with MTOM dependency, Rec20/22 - wait to close next week, also
evaluate Yves binding framework, are there any diffs between Rec20/22 (and maybe
458)
 MarkN media type registrations 
 LC schedule (6 weeks behind)
 Issue 447 - realizing MTOM in HTTP bindings
 Issues 465/466
 Close Rec 18/19/23 - we have proposed resolutions for these

 
 
        

3. Approval of minutes. Minutes just posted, not read, minutes approval hereby

postponed to next week



        

4. AI Review

  Status change - Noahs item is done
  Canonical base64 text is done
  Also Done:
     Yves message to WSD
     Yves Rec20/22 
  Mark 464 was waiting for new version of RepHeader
  Mark has dup AIs
  Gudge - did he send email regarding base64 to Schema, NO, will do and send
  MikeM - resolving some editorial URI issue regarding bookmarks


        
        

5. Status

LC Status
  The schedule we estimated in Cannes would have us in LC this week
        What is left to do:
   issues 465,466, - today
   rec22, 458 - almost there today, by next week
   447 dependent on 1.1 analysis, today or next week
   443 media typing Anish - be deferred till media type doc is stable, now is stable - so next week
   464 did we meet our requirements
   2 pieces of text Herve is working
  Hence by next call, all issues should be resolved
  Another draft of docs by late next week
  Week of May 3rd-10th to resolve any issues
  Week of 17th, generate the last call docs
  LC then May24th.
Jacek - 443 may be also dependent on WSD work, XOP sect 3.3 would have to be
removed/modified
David - OK we will figure out how to do that
Further ahead:
  Originally we had CR in June and PR in August
  So now, CR in July, Aug is prep for PR, Rec is for Oct
3 weeks between LC and CR (originally) -- cant shrink this time

Media Type Reg
MarkN - 2 Internet drafts, application/soap+xml major change is only XML 1.0 from
0.4 draft, this is in section 2 (last paragraph) and where ever XML was noted.
Feedback that this is only in Intro, but not is encoding considerations. Mark
reasons why this is OK. 

Other draft is priliminary draft for XOP media type, application/soap_xop+xml.
This just identifies the root part, and not the package too much. Not ready
until we have stable XOP draft. LC is sufficiently stable. Please everyone read
the subtle language Mark constructed.

Include/XInclude FAQ - make an Editors Copy. 
David - no reply from Core, next week decide whether to make it a WG note or not.
Will inform Core unless they tell us otherwise. Email WG with final draft.

XMLP/WSD Task Force and WSDL Media Type document,
David: Anish is not here. Is anyone here else following this? 
MarkN: Conf call monday, the status is that the state is good and it is being
written up currently. 
Gudge: True
David: We need to push on this, XOP section 3.3 is dependent on this document

Discussion with DOM WG regarding value/lexical spaces, APIs. What do we want to
ask them? See http://lists.w3.org/Archives/Member/w3c-xml-protocol-wg/2004Apr/0021.html
MarkN: For implementations to take advantage of what we give them - they need
access to the raw binary data. Hence the interface needs to change, without going
through the encode/decode loop. DOM group we can influence directly. SAX project
also. We should draft email to DOM to suggest a new requirement. 
Gudge: Thinks DOM WG is going away
Yves: Should close at some point - but will need to do maintanance
David: Need a volunteer to write email
MarkN: I will do it. Suggest an approach to DOM also. We need to make sure that
other W3C groups support our work

MarcH: JAXRPC 2.0 jsr224 considers MTOM/XOP on its plate, no work in JAXP though
(responding to MarkN)



        

6. Attachements

- Start with Noah's text rec20 and rec22
All: reading Noah's text, deciphering notation
Yves: Section 5, paragraph 4 - only CII that is whitespace needs to be handled. 
Gudge: Even the new Infoset uses the old NEL definition. The NEL codepoint is
not in the Infoset
David: Issue from that para: EIIs can have CII that are XML 1.0
Gudge: XML 1.1 parsers will translate to a 0#a, 1.0 parsers will blow up
Yves: In the infoset rec they reference 1.0 and 1.1 to handle this
David: any other issues with Noah's text. NO issue. This text will be part of
the rec20 resolution package.

Working Group agrees that Noah's text is accepted for SOAP 1.2 errata.

***   Jacek signs off for last time :(   ***

Noah's Unresolved Issue, should we specify a standard fault?
Gudge: define a fault subcode - unsupported XML version. Can we add that in an errata.
David: it is more than editorial - this changes how the processor works
Gudge: redefining the fault - thinks do not bother
Yvess: The error can be covered by the existing unsupported media type fault.
David: Should we change the text around unsupported media type?
Gudge: no

WG decision - will not create a new fault, the unsupported media type will cover
this - as it is specific to XML 1.0

Impact of Binding Framework Analysis
Yves: every binding is not required to use XML 1.0 serialization
Yves: only reference to serialization - deferred to media types
Yves: m-t doc states xml 1.0 serialization
Yves: part1 soap binding - nothing stated explicitly how binding is serialized,
so we can default it to 1.0
David: so do we need text in 7.4?
Yves: maybe also in the primer
David: yes, we have to discuss primer
Gudge: +1 to Yves evaluation
David: Yves can you draft clarifying text for next week
Yves: yes I can

David - last item is Yves's evaluation of whether rec 20 and 22 are the same
Yves: in last case, they are abit different, but we are covering them at the same time
David: So the full package we are creating will cover both Rec20/22?
Yves: Yes

David checks that Herve will generate MTOM text from Noah's SOAP 1.2 Errata text
Herve: Yes

458 - Impact of XML 1.1 
David: is this the same as 20/22?
David: Yves will the resolutions to 20/22 also cover 458?
Yves: we can try
David reads 458, seeing if there is anything we have not covered ...
reads through the issue
WG: agrees without objection that all the issues raised in 458 are duplicates
of issues raised in 20/22
David: Issue 458 is closed

David: Issue 447
Gudge: Doesn't Marks media type handle this issue
MarkH, David: yup
Gudge: we can close this, keys off the mtom media type
Gudge: Proposal - SOAP HTTP implementation will use presence of mtom media type,
to determine whether to do mtom processing
WG agrees without objection to close the issue with this proposal.

David: 465/466
David: 465 - spelling error or new term. Text referenced in issue no longer exists.
Chair proposes to close this issue without action.
WG agrees without objection to close the issue. 
David: Tony Graham is nominated for closing text. 

David: 466
Gudge: the restriction no longer exists in XOP, still exists lower down in the stack
David: we propose to close 466 taking no action because the restriction is not in
XOP, but in the binding, and people are free to create own binding that does not
have such restrictions.
WG: No objections to closing issue with this proposal
David: Issue 466 is closed
Gudge to send closing text and email.


David: next week we should close Rec issues 20/22 package, and issues 443, 464,
and then we are close to being done