W3C XML Protocol Working Group teleconference, 18 September 2002

1. Roll call

Present 23/19 Excused Regrets Absent

2. Agenda review, and AOB


3. Approval of 11 Sept telcon minutes: postponed


4. Review action items


5. Status reports

5.1 Primer
  Almost everything is done. There is only one that is not done yet.
  That is the extra text that Noah had once written on intermediaries.

5.2 Spec
  We got a bunch of the closing text.

5.3 LC Issue List

5.4 Implementation tracking
  Nothing to report

5.5 Media type draft, IANA application
  Nothing to report

5.6 Planning Oct/Nov f2f
  A bunch of responses by email. Currently we have to pick the later
  October date (29-31).
  Is there anybody who objects to the first question? ... None.

  Is anybody not able to attend the earlier days? ... None
  Is anybody not able to attend the later days? ... None

  At the moment, we will be headed for the later date (29-31 October).

  Is there anybody who probably could host on 29-31 on the east coast? ...


6. Attachment Feature document

  Any comments to the content? ... No discussion
  Any objection to sending the current copy of the Attachment Feature WD
  to W3C for Last Call? ... No objection

  No objection from the WG to as taking the current version of Editor's
  copy of the Attachment Feature document to Last Call.


7. LC Issues

-- Issue 338
(Asir) Issue 338 has two parts. Both parts are about the state machine.
       We decided not to take any action and close the issue.
       I was able to pick up the rationale for Part2 from the minute,
       but I was not able to come up with the rationale for closing
       Part1 of the issue.
(Chair) OK. And you are still searching for the rationale as to why we
        decided to keep it separate?
(Asir) Yes. I think we talked about the streaming and timing issues.
       But I couldn't pick up the rationale from the minute.
(Chair) Is there anyone on the call who can regenerate the rationale
        for Asir so that he can complete his action on 338? ... none
(Chair) Asir, could you send an email to the member list describing
        your problem here and the rationale you are seeking?
(Asir) OK.

-- Part 2, Table 17 discrepancy
Marc outlined his thought about the changes in table 23 or 17... .
(Marc) Options are we can change the detail in table 17 such that 400.
       I'm not sure we need to change the entries 401, 405, and 415.
Discussion between Marc and Henrik.
(Chair) Herve will re-modify the proposal.
(Jean-Jacques) The last part of 167 in September xml-dist-app.
(Marc) I wonder if one simplification ... just pulled out particular
       transition at before the table to make it clear that any of
       these code to potentially come back with SOAP envelope in the body.
(Chair) I think we need the text on this. Marc, would you privide the text?
(Marc) Sure.

-- Issue 292, how to combine RPC and encoding faults
(Chair) Noah provided the text in xml-dist-app/2002Sept/0173.html .
(Chair) I think that consensus has reached around about 173 maybe it's 175.
        Jacek had a question about the terminology around "MUST
(Jacek) I did see the final proposal. My original problem was that
        the text "a fault MUST be generated" ...
(Chair) Are you happy with the 173 text?
(Jacek) Yes. 175.
(Chair) The text in the 175 is the final...
(Chair) Issue 292 is closed with the text which is within the <proposed>

-- Issue 250, definition of a role (editorial)
(Noah) This is the question whether we properly define the special
       roll name like 'next'... .  I sent a response to Joseph basically
       explaining we tweaked up the text. I think he is questioning
       the decision made by the WG not my explanation... .
(Henrik) I read the response also and had the same reaction. I suggest
         to try once more.
(Chair) We re-explain to Joseph.

-- Issue 297/185, generic types unnecessary.
Asir explained the rationale.
(Jacek) I disagree with most of Asir's position. I think removing generics
        is the better way to go.
Other comments from Henrik and Noah.
(Asir) We concur the majority.
(Chair) Does anyone object to removing generics? -> no objection
        Issue 297 is closed by removing generic types from the specs.

-- Issue 302, data model edges orignating and not terminating
(Jacek) I think the choice is between the copy and the alternate proposal.
        The editor's copy is acceptable.
Discussion between Noah and Jacek.
(Chair) In bullet 4 there's a mention xsi:nil and is not highlighted.
        I think we need to see the text back again.
(Chair) I'm getting very nervous with putting proposal of undecided things
        in the editor's copy. It is not a good place.
(Chair) Henrik, would you take an action with Gudge to get those?
(Henrik) Yes.

-- Issue 231, what is the difference between a struct and an array in the
edge case?
Jacek explained the options he provided.

straw poll
status quo: 3
add a new attribute: 4
other TBD mechanism: 0

Anyone object to status quo? ... None
Anyone object to adding a new attribute? ... None

(Chair) Vote between status quo and a new attribute. We will do this by
        simple majority. We should do this by member company.

Status quo: none
Adding a new attribute: Systinet, AT&T, Sun, SoftwareAG, IBM, Oracle, Canon

(Chair) We will create a new attribute.
(Jacek) I can write the proposal text for the new attribute.

Meeting adjourned.