W3C XML Protocol Working Group teleconference, 1 May 2002

1. Roll call, scribes for minutes/action items

Present 24/20 Excused Regrets Absent

2. Agenda review, and AOB

DavidF went through agenda.

I18n charmod. DavidF suggest that we ask XMLP WG reviewers of previous
Charmod spec to look at I18N's resolution of our comments. Paul Denning
(a previous reviewer) agreed to figure out what happened to our comments.

3. Approval of March 27, April 17, April 24 telcon minutes [see member webpage]

No objections to approving all minutes as they are posted. They are approved.

4. Review action items, see [1]


* 2002/02/26: Editors 
Insert the notion of "root" in the data model 
DF asks the WG whether or not to remove "root" attribute and the possibility of
top-level multiref?

Some discussion. Not clear whether it is orthogonal to top-level multi-ref or not.
HFN: proposal is to remove the root attribute.
NM: points out that we don't actually prohibit multi-ref.
MH: points out that SOAP 1.2 doesn't really say either way.
NM: It depends on how encodingstyle attribute is defined
SW: notes that it is not clear what the choice are
NM: suggests that we postpone the discussion to after discussion on 194 and 195

* 2002/02/25: TBTF 
Resolve ednotes in 6.1.5 and 6.1.6

These are not the correct section numbers. TBTF has dealt with this -
report is coming later. Sense is that they are editorial.

5. Status reports

In general, reporters should be able to describe what is still on 
their to-do lists, and what is their plan for completing those to-do's.
-- Primer
Nilo not here. DF believes the document is up to date

-- Spec
Few items still on the todo list (some came up today).
Readability editorial review has been incorporated.
Expects 2 days for incorporating rest of todo items
and one more day for TBTF asynch request/response proposal (if adopted).

-- TBTF
SW: Sent status report to mailing list. Some was about old feedback
items from December. Henrik started discussion on one, another one on
HTTP 202/204 still missing. Have a few editorial edits (part of the 3
day editors' estimate). Question outstanding regarding missing data in
input column in HTTP state transition table.
HFN: Proposal: follow-up on the mails and see what they say and if more
discussion then log an issue.
CF: Purpose is also to ask for additional input from the issue raiser.
We are still discussing issue 202.
DF: Put date in for feedback deadline would be a good idea. Set up new
TBTF call.

-- Conformance work, see plan [2] and report from 29 April telcons. In
particular, what has the conformance team learned in the last  few days, and
what is their status with regard the plan outlined in [2].

Anish: (referring to email laying out tasks) - we are one day behind schedule:
  Task a) is done except for 5 assertions (81 tests in all)
  Task b) (assertions for part 2) done. 
  Task c) not done
  Task d) not done
DF: timetable was vague for c) and d) because there was no sizing for task (c).
Richard, you took the task of sizing c), what is it?
Richard: Part 2 tests: 62 assertions of which no tests are written
Anish: the exact number is debatable
DF: we have to evaluate all the assertions anyway, so we say on the order of
50 tests for task c)
HFN: When are we going to have a complete draft?
DF: The schedule says next Thursday we will have a complete draft, is that correct?
Anish: yes it does say that
HFN: W3C moratorium for publishing May 1 to May 14
DF: Can we have a complete document by 14th?
Anish: We can provide a complete document including task c) BUT NOT d)
by May 14.
DF: I note there are only two people working on the document, the tests, etc. I
think d) be worked in parallel, so who can volunteer for d) in the next 7-10 days?
CF: If we had a diff between the snapshots then I would be glad to do it on a
weekend
JI: Also has time to look at it. Output is a set of actions to perform
on the test document
DF: (To editors) Is there any markup in the spec that delineates assertions?
MH: No, this was not done.
HFN: So, *can* have a complete up to date draft by May 14?
Anish: We can certainly shoot for it
HFN: Yes/no?
Anish: no
Anish: We can commit to May 21.

-- Usage Scenarios
JI: Nothing to report
DF: anything to do to create a WD from the document?
JI: I'll take a look and report
CF: What's the interaction with WS-arch WG?
JI: We have come to a level of completion. David Orchard will take that
and move forward in the context of WS-arch. 

-- Requirements document
Bob: Nothing to report
DF: anything to do to create a WD from the document?
Bob: Nothing as far as I know
CF: In terms of publishing as WDs, the pub rules have changed
DF: yes, but we need to stabilize the content of the documents first

-- Email binding
HM: Depends on req/res statemachine discussion (see below) - some
changes may occur. If no change to state machine then no change.
DF: How much work if these changes are adopted?
HM: roughly a day

-- Readability edits
DF: completed the task and sent result Monday

6. Loose ends? E.g. are there any to-do's, issues etc that we have not captured?

None mentioned.

7. LC planning

During last week's telcon, we discussed going to Last Call 
with and without a Test/Conformance document. We do not have to go to Last Call 
with this document, however, given that we are planning to skip CR we 
will certainly need such a document sometime during Last Call. The issue then seems to
center around whether or not we want to go to Last Call with all our
documents at once, or whether we are comfortable with one arriving shortly
after the Last Call announce. One factor to consider here is 
that the W3C has announced a publication moratorium starting May 1 and continuing
through May 14.

Also, we previously agreed [3] that we would make a go/no-go LC decision at
this telcon (modulo giving our European colleagues a chance to make their
voices heard in such an important decision). We clearly have a number of
outstanding to-do's, and so we need to decide the threshold at which we can
declare each one complete enough to go to Last Call.

The Chair will post a summary proposal for the disposition of our various documents
in Last Call before the telcon.

DF: The go/no-go call had the assumption that we had all issues done.
This is not the case. Two issues seem key: a) moratorium until
May 14, b) conformance/test document is still not ready. Regarding the latter, are
we prepared to go to LC without the document? Checked with W3C, it's
ok to delay test document 1-2 weeks after going to last call on the
spec. However, it does affect that we are not planning for CR - seems
like a missing piece. There has been a lot of interest in going to LC
sooner rather than later.
HFN: It is important that we can trust the test conformance document
moving forward. Can we have the test editors report progress every 4
days or so?
NM: When do you expect that we can have a clean draft that we can review.
DF: Around the 10th?
NM: Some people are at the WWW2002, how does this affect review time? A
week is about the right time.
SW: More comfortable with a 5 day review period.
MH: Some TBTF issues are relative big in terms of what they touch. Let us
first see what the WG's decision is with regard these issues.
DF: OK, we'll move discussion on timing to the end of the call.

8. Remaining active issues

-- Review and decision on whether or not to accept TBTF proposal for
supporting overlapping requests/responses within the single-request-response MEP [7]

MH: No connectivity
SW: Walks through summary of TBTF on the issue of whether we allow
streaming or not and if so then how we affect the state machines
NM: Summarizes http://lists.w3.org/Archives/Public/xml-dist-app/2002Apr/0327.html
DF: Any objection to incorporate MH's proposal into part 2?
CF: How can MUST checking be done?
NH: Just say that existing processing has to be followed
HM: How long will this take to put into the spec?
HFN: 1 day

DF: Regarding timing of edits, review etc. Two ways of moving forward have been
suggested. One way is to conduct a major review very soon (Saturday?), and then
conduct a quick review on the diff when we reach LC. A second way is to wait until
we have what we consider to be the final LC WD, and conduct a major review on that
document. I am concerned about the former approach because experience suggests the
diff may not be so small and easy to reconcile as we might now believe.

Other WG participants agree.

DF: We will not start a major review Saturday but I will consult with editors
to find a date.

Meeting adjourned

References.
[1] http://www.w3.org/2000/xp/Group/Admin/#pending
[2] http://lists.w3.org/Archives/Member/w3c-xml-protocol-wg/2002Apr/0180.html
[3] http://lists.w3.org/Archives/Member/w3c-xml-protocol-wg/2002Apr/0067.html
[4] http://lists.w3.org/Archives/Public/xml-dist-app/2002Apr/0234.html
[5] http://lists.w3.org/Archives/Member/w3c-xml-protocol-wg/2002Feb/0161.html
[6] http://lists.w3.org/Archives/Public/xml-dist-app/2002Apr/0150.html
[7] http://lists.w3.org/Archives/Public/xml-dist-app/2002Apr/0395.html