W3C XML Protocol Working Group teleconference, 21 November 2001

Minutes of XML Protocol Working Group telcon, 21 November 2001.

1. Attendance

Present 21/18 Excused Regrets Absent

2. Review of agenda and AOB

No changes

3. Minutes of Nov 7th and Nov 14th, both approved as posted (including

HugoH's changes to Nov 7).

4. Review Action Items

5. Status reports

* f2f
Agenda stable; some minor revisions will be posted today/tomorrow to
incorporate issues left over from today's telcon and ETF proposals.

Mario, MartinG and Highland intend to dial-in. Hugo to arrange for use of
W3C bridge with ~6 lines.

JJ wonders how much interaction/participation DF expects from remote
attendees?
DF: always an experiment to see how this works. DF will do his best to be
inclusive of remote attendees during discussion. PC suggests DF annoints an
advocate for remote attendees that express concerns via IRC. DF suggests
that remote attendees use IRC during f2f. CF or MH to be online during
meeting to catch IRC traffic from remote attendees. Hugo to post details of
W3C bridge dial-in to list.

* Primer
Nilo has published substantially revised version which is on the f2f
agenda, he hopes we have time to review and send feedback/comments.

* Editors
Nothing to report. DF calls editors attention to f2f agenda item that
contains multiple editorial issues.

* TBTF
Have been working on Introduction document (for section 5 of SOAP 1.2 Part
1) which has been made available to WG and on agenda for f2f discussion.
Presentation by TBTF at f2f.

* ETF
Continues to go through issues. A couple proposed resolutions on agenda
today. 4 issues outstanding... getting close to the end of their work.

* Conformance
Hugo has updated the collection document with assertions found in editors
copy of spec. There is an agenda item for f2f; encourages f2f participants
to bring tests. The conformance document is up to date and in synch with
the spec.
Suggests that spec editors co-ordinate with conformance team to point out
any changes in the spec.

* Usage Scenarios
A new version has been published, and is on the f2f agenda.

* SOAP with Attachments
DF: ball in his court to draft S+A text for Charter. Paul Grosso (CG) has
pointed out an "XML packaging" proposal, wondering if that is of interest.
DF will forward PG's email to XMLP (member) list.

6. Issue #32

SW has proposed a resolution to close 32. No discussion on the list about
it.
The proposed resolution was accepted by the WG without objection.
SW to send email to xmlp-comment.

7. Issue #172

NoahM sent a summary email containing proposed resolution text. It was
proposed to close issue #172 with the text in that email (as proposed by
MarcH).
Proposal accepted without objection.
MarcH to send email to xmlp-comment.

8. Issue 146

SW discusses his revision text proposed in [1]. Should we close issue on
special status of Body first? Nobody raises any serious objections.
Proposal to close issue 146 with SW's text in [5] (last paragraph).
Proposal accepted without objection.
SW to send email to xmlp-comment.

9. Issue 130

Discussion around Gudge's proposal. DF suggests we tackle heirarchical
versus flat and then address element v attribute issues. HFN also discusses
the problem that only one fault code can be represented. HFN suggests that
fault mechanism should be extensible. Discussion around distinction between
<a/><b/><c/> vs <a><b><c/></b></a> HFN suggests that "flat" would be
equivalent to a stack trace. HFN wants to apply KISS principle. Seem we
have two styles of preference; lexical versus semantic .... Gudge suggests
that he create examples of single fault, single fault with refinement, etc.
with flat versus heirarchical. HFN suggests we can remove the dot notation
entirely. DF suggests one way to deal with this would be to write up a
proposal askiing feedback from dist-app and SOAPBuilders on the options of
(a) removing dot notation altogether, and (b) adopting a hierarchical
notation.
MH points out that Gudges proposal also addresses issue #143, changing
client/server to sender/receiver.
CF proposes to close issue 130 by removing dot notation mechanism for
faults codes from spec, and insert an Editors Note noting the removal and
saying the mechanism will remain removed unless there is compelling
feedback to reinstate it.
Proposal was accepted without objection.
MartinG to send note to xmlp-comment.

HFN suggests that the discussion around 130 has raised a new issue; we only
have one high-level status code (e.g. "Client", "Server"). HFN to write up
proposed new issue and send to list.

10. Issue 101

Review Noah's proposal referenced in agenda. All but Jean-JacquesM agree
that the Body is NOT symmetrical with header blocks. JJM: asymmetry may be
OK if there can be more than one Body. HFN: two issues, what is main intent
of message and also from processing POV what are the differences? Much of
the complexity that the changes add to the processing model is that it
keeps some notion of body blocks when we actually just have "the body". CF:
ideally, headers and body would be symmetrical but in reality (and in the
spec) they are not, and we should not try to pretend that they are.
Discussion to be continued by email and during f2f.

Meeting adjourned.

[1] http://lists.w3.org/Archives/Public/xml-dist-app/2001Nov/0220.html