W3C XML Protocol Working Group teleconference, 4 September 2002

Minutes of the XML Protocol Working Group telcon, 4 September 2002

See also the IRC Log

1. Roll call

Present 24/19 Excused Regrets Absent

2. Agenda review


3. Approval of f2f minutes:

No objection to the approval of the posted minutes.


4. Review action items

4.1 [2002/07/31: Asir] Draft email to xmlp-comments to close
issue 338 with status quo: remains pending. 
4.2 [2002/08/01: Editors] Make the proposed change wrt issue 192 
resolution: still pending. 
4.3 [Gudge] Research infoset properties of header and body children, 
and suggest appropriate changes: still pending. 
4.4 [Editors] Clarify the relations between code, subcode and detail
 element, but at least 1/3 of the size of the proposed solution for 
issue 320 resolution: pending. David will put it on next week's agenda 
4.5 [2002/08/01: Editors] fix all issues marked 'Editorial' in the 
issues list: closed. 
4.6 [2002/08/01: Yves/Carine] Investigate what more can be done under 
the existing charter of XMLP in the time extension (post REC): pending 
4.7 [2002/08/01: Editors] make changes as in the proposal: pending 
4.8 [2002/08/02: Noah] Issue 292 -- propose a resolution that the
processor may elect to report any fault it chooses (when multiple faults
 exist): discussion already started on the list; nothing has been done,
yet. Action closed! Will be on the list for next week.


5. Status reports

5.1 Media type draft, IANA application Mark Baker resigned
from the group, but is happy to help to take the IANA application forward
in a public manner. Two Options: 1) send in a complete internet draft now
including a note saying the technical content will eventually be contained
in the SOAP 1.2 Rec. 2) send in a "shell" internet-draft pointing to the
next SOAP 1.2 spec. Anybody interested in this topic is encouraged to
contact the chair directly.

5.2 Primer (Nilo) Working on the comments received on the primer. Very
large number of them incorporated. The editor's copy of the updated last
call WD will be made available on the public page. (The updated versions
of the other parts of the spec are publicly available already.)

5.3 Spec (Martin Gudgin) Every issue that was resolved at the f2f and
every editorial issue on the issue list has been incorporated with
exception of issue 173. Diff markup complete in part1, part2 not finished
(Noah) For a number of editorial issues the editors decided to do
nothing. The originators of these issues are probably waiting for some
feedback. Closing remarks should be sent to xmlp-issues also for these
issues which were not handled by changing anything. 
(David) One big e-mail would be sufficient for every class (change, do 
nothing, etc.) of editorial issue closed by the editors. Editors should 
respond at more length to comments if they think it is appropriate.

5.4 LC Issue List (Yves and Carine) The issue list is up to date and is
not expected to grow in size.


6. Attachment Feature document

(Chair) Question to WG: Should the current document be sent
forward as a LC WD? 
(Marc) Just prior to the meeting I sent some comments to the public
list. Some of these were relatively trivial editorial things but there
was one substantive question around the inclusion of a basic attachment
referencing scheme like CID. I also note that there are comments from  
Chris Ferris that have not yet been addressed. I would like for Chris  
and my comments to be addressed before we decide on proceeding to last 
call or not.
(Noah) Reluctant to include implementation specific stuff like CIDs. 
(Chair) Proposal: Let some discussion happen on DistApp
and then revisit this question next week.


7. Implementation tracking

(Jacek) Thinks that all discussion issues on the list of
implementation features have been resolved.

8. LC Issues

(Chair) Regarding LC issues, there is a very high bar to changing the spec.
(Henrik) That includes removing things from the spec

- Issue 221 Is the working group prepared to accept the revision 
(Gudge) Can live with Noah's (reference 8 in the agenda) proposal. 
(Henrik) Can also live with ref 8, although I do have some concerns. 
(Marc) This seems to be an endless issue ... I propose we move forward. 
(Chair) Are there any objections to amending the existing resolution of the 
issue with the proposed revision? No objections expressed. The 
revision is accepted.

- Issue 227 (Chair) Are there any objections to amending the resolution
record (in the issue list) as proposed? No objection expressed. The
amendment to the record of the resolution is accepted.

- Handling of any closed issues for which we have received pushback
No such issues have been identified.

- Issue 303, "Fault for broken array attributes?" 
(Jacek) Concern is just for consistency as already stated within several 
e-mails. Prefers removing the attribute, but do nothing would also be a ok with him. 
(Chair) Is there any objection to closing issue 303 by accepting the status 
quo (i.e.  doing nothing)? No objection to so closing the issue. The issue is 
closed without action.

- Issue 234, "change max arraySize from '*' to "unbounded" for consistency
reasons w.r.t. XSD" 
(Chair) WG commentators to date on this issue have wanted to maintain the status 
quo. One said that it is not necessary to bring schema and instance documents 
into alignment because they are different. 
(Chair) Is there any objection to closing issue 234 by
accepting the status quo (i.e. doing nothing)? No objection to so closing
the issue. The issue is closed without action.

- Issue 325, "define an xml schema based encoding" 
(Anish) as originator and based on a further close reading of the SOAP 1.2 
spec, I believe there is actually no need for such an encoding 
(Chair) Is there any objection to closing issue 325 by accepting the status 
quo (i.e. doing nothing)? No objection to so closing the issue. The issue is 
closed without action.

- Issue 297, "generic compound types unnecessary" 
(Gudge) Notes that this is the same as issue 185 which we settled before by 
asking for feedback through an ed note in the spec. What was the feedback 
received on this issue? Feedback: 2 for removing it, one for keeping it 
(Jacek) Point against generics: they bring the power of XML into SOAP encoding 
and thus complicate it. 
(Gudge) It should not be too disruptive to remove generics
(DavidO) +1 
(Camilo) We are very interested in keeping generics because
we're using them already. 
(Chair) Suggestion to postpone the decision, until we have a rationale from 
Camilo/Asir for keeping generics. WG will discuss again next week.

- Issue 302, "re. data model edges originating and not terminating at
(Gudge) This is just an editorial issue. 
(Jacek) Will leave it to Gudge to create a proposal and then decide if is 
editorial or not. 
(Marc) Has it to do with handling dangling references instead of handling nulls?
(Jacek) This is not the same issue. 
(Chair) We'll wait to see some text from Gudge.

- Issue 231, "what is the difference between a struct and an array in the
edge case?" 
(Noah) Should we state in the case where you have an empty
construct and no item type and no size this is a generic? How to
distinguish empty arrays from empty structs? (?) Can an array have
different types in its children? 
(Jacek) arraySize should be mandatory.
Necessity of a SOAP data model schema language, currently often WSDL is
used to come to an decision which boils again down to the usage of a
schema language. 
(DavidF) Does the resolution of this issue change if we
decide to remove generics? 
(Chair) It would be helpful the have a summary
of the various options.

- Issue 306, "is use of Appendix A optional?" 
(DavidF) Is one obligated to use the Appendix A mapping in RPC? 
(Henrik) (BTW, appendix A is now called appendix B) Appendix B is normative. 
If one uses name in the RPC names which cannot be represented directly in XML, 
it is a good idea to tell them to use appendix B. Reference to appendix should 
be in sect. 2.1.2.
(Chair) straw poll: appendix B must be used: 1, optional: 3 Discussion
will be put to E-Mail thread.