ISO/ IEC JTC1/SC22/WG21 N0733

                                     Document Number: WG21/N0733
                                                Date: 11 July 1995
                                             Project: Programming Language C++
                                            Reply to: Dan Saks
                     ISO/IEC JTC1/SC22/WG21 Meeting No. 13
                               9 - 14 July 1995
Monterey Marriott
350 Calle Principal
Monterey, CA 93940 USA
1     Opening and introductions
      Harbison convened the meeting at 18:00 (PDT) on Sunday, 9 July 1995.
      Saks was the secretary.
1.1   Welcome from host
1.2   Roll call of technical experts
      The attendance list appears as Appendix A.
1.3   Select meeting chair
      Harbison offered to chair the meeting.
1.4   Select meeting secretary
      Harbison affirmed that Saks is the secretary for WG21.
1.5   Select language
      Harbison suggested that WG21 conduct this meeting in English, and no one
1.6   Select drafting committee
      Saks (as secretary) is a standing member of the drafting committee.
      Harbison asked others to serve on the drafting committee for this meet-
      ing, including at least one WG21 member whose native language is not
      English.  As usual, Corfield, Hartinger, and Unruh volunteered.
1.7   Adopt agenda
      Harbison submitted MT-2 as the meeting agenda.  Jones suggested moving
      item 2.3 to 1.8.1.  WG21 accepted the agenda with this change.
1.7.1 Recognize documents
      Harbison introduced:
      --  MT-1: The Delay in Submitting the CD, and What to Do About It
      --  MT-2: WG21 Agenda for Sunday, July 9, 1995
1.8   Approve minutes from previous meeting
      Saks submitted N0660 = 95-0060 for approval as the minutes of the pre-
      vious meeting, which WG21 accepted.
1.8.1 Review action items from previous meeting
2     Status, liaison and action item reports
2.1   Small group status reports
2.2   Liaison Reports
2.2.1 SC22 plenary report
      No report.
2.2.2 SC22/WG11 (Binding Techniques) report
      No report.
2.2.3 SC22/WG14 (C) report
      Deferred to the joint meeting with X3J16.
2.2.4 SC22/WG15 (POSIX) report
      No report.
2.2.5 SC22/WG20 (Internationalization) report
      No report.
2.3   Review action items from previous meetings
      No items from previous meeting.
3     New Business
3.1   Review changes to working paper
      Deferred to the joint meeting with X3J16.
3.2   Future meetings outside the US
      Corfield offered Programming Research and the BSI as hosts for the
      meeting in July, 1997 at London, UK.  Harbison suggested delaying the
      meeting one week from what would be the normal meeting date so that
      members won't have to travel on the July 4th weekend.
3.3   Review CD submission delay
      Bruck said the standardization process and the relationship between WG21
      and X3J16 are unclear.  He disagreed with Harbison's decision not to in-
      form other members of WG21 of the disputes described in part 1 of MT-1,
      particularly because procedural objections might have been raised with
      SC22.  Harbison defended his decision.  He said all the parties in the
      disputes ultimately resolved their differences among themselves to
      everyone's satisfaction, so it indeed appears that there was no need to
      involve others.
      Koenig stated he did not think all parties were satisfied with the
      resolutions to the disputes.  In particular, he was not.  He believed
      that Harbison told him that he had to resolve all library issues in
      favor of Plauger, or there would be no CD.  Koenig agreed with Bruck
      that the fundamental problem is that we don't really have agreement on
      the procedures for drafting the document.
      Plauger supported Harbison's approach to resolving the disputes but
      suggested that Harbison should have informed WG21 delegates sooner as a
      matter of courtesy.
      Harbison asked WG21 to consider repealing the "Skaller resolution",
      which Plum summarized as: WG21 has empowered the editor to make sub-
      stantive changes to the draft to correct flaws provided (1) that the
      changes do not contradict WG resolutions and (2) that the editor
      documents the changes using editorial notes in the draft.  Plum went on
      the say that, although the editor can delegate editing work, the Skaller
      resolution did not delegate the authority to make unapproved changes.
      Koenig said the scale of the editing chores is too much to get right in
      the alotted time.  Harbison suggested doing more work in post-meeting
      editing sessions.  Koenig countered that some members can't attend such
      meetings.  Harbison said we need to find members who can attend to whom
      we can delegate editing responsibility.
      Lajoie opposed repealing the Skaller resolution.  Koenig insisted that
      the editor needs final authority to decide what goes in the document.
      Plum said, and Unruh echoed, that the concern is not over changes and
      editorial boxes in a draft that stays within WG21, but rather with
      changes in a draft that goes (via the convener) to SC22 before all
      committee members have had a chance to review it.  Plum expressed his
      opinion that those drafts must embody WG21 decisions.
      Stroustrup also opposed repealing the Skaller resolution.  Repealing the
      resolution would reduce the chance of making mistakes, but it would also
      reduce the amount of work accomplished.  He suggested that, rather than
      repeal the resolution, we should make it clear that anyone reviewing the
      draft can insist on adding an editorial box noting his/her concerns.
      Harbison asked if WG21 should change its editing procedures.  Much more
      discussion followed.  Harbison said he heard a clear consensus that WG21
      needs more traceability for changes in the draft.  Lajoie suggested that
      a change where "we all believe the committee agrees on this change" need
      not be documented.  Others murmured disagreement.  Plum wanted to make
      sure Stroustrup's point wasn't lost: any participant in the review
      process can get a concern raised to level of an editorial box.
      Harbison summarized the outcome of the discussion:
      --  The Skaller resolution shall remain in effect.
      --  We must do better at tracing changes to the draft.
      --  The editor has the final say about what goes into a draft, but any
          reviewer can request an editorial box.
      --  The convener shall ensure that the agreed upon process is followed.
      Lenkov said that the Library working group (WG) needs to draft its re-
      solutions with more precise wording.  Harbison also said that repre-
      sentatives of the Library WG should be present at post-meeting editing
      sessions.  Stroustrup agreed to discuss these concerns with the Library
      Harbison said he'd try to write a description of the editing procedures.
3.4   Review delegation positions on CD
      Hartinger said it appears that, barring surprises, Germany will vote
      "yes with comments" on the CD ballot.  Bruck said Sweden will most
      likely vote "yes with comments".  Santo said the same for Japan.
      Plum said that, over the weekend, X3J16 members looked at 195 pages of
      comments from the US public review.  He guessed that the US will vote
      "yes with comments".  Lajoie said Canada will not conduct a public
      review.  She expected Canada to vote "yes with comments".
      Corfield said the UK are accumulating a wealth of comments on the
      draft.  They will likely vote "no with comments".  He expected the UK to
      collect around 1000 different issues regarding the CD.
3.5   Schedule options
      Harbison observed that WG21 has had a problem "turning around" the draft
      document within five weeks after a meeting, which we tried to do more
      than once.  The document for next stage, DIS, must be a final draft and
      shouldn't be rushed.  He suggested using the November 1995 meeting to
      resolve NB (national body) comments on the CD ballot, and using the
      March 1996 meeting for final wordsmithing (which we would need whether
      we advance to DIS or submit another CD).
      Harbison noted that SC22 changed the closing date for the CD ballot from
      August 28 to September 28.  This means we cannot include the CD ballot
      results in the pre-Tokyo meeting mailing, which will affect our ability
      to resolve CD ballot comments in Tokyo.
      Koenig expected to see feature changes in the draft at least through
      March 1996.  We won't know if this will be the case until we see all the
      comments from the CD ballot.  Stroustrup thought we would only be fixing
      problems after this meeting, not adding features.
      Harbison raised the possibility that we might need another CD ballot,
      and might wish to plan for it now.  Plum lobbied against another CD.  If
      there are problems, let's fix them.  If we were to go out for another
      CD, it will only confuse the world.  Lajoie agreed.
      Plauger explained that, if there are too many "no" votes or too many
      comments on the CD ballot, the SC22 Secretariat will probably be in-
      clined to stage another CD ballot even if WG21 recommends going to DIS.
      Harbison said the consensus seems to be that we should wait until the
      March 1996 meeting to release a new draft to SC22.  Whether we submit
      that draft for a second CD ballot or a DIS ballot, or whether we're even
      done resolving comments, depends on our progress by then.  Stroustrup
      said he thought SC22 would prefer we fixed all the problems before
      resubmitting the draft.
3.6   New business
      Lenkov informed WG21 that he will be resigning as chair of X3J16.  He
      will not be chairing the joint sessions on Thursday and Friday of this
      meeting.  Clamage will submit his name to X3 to serve as the new chair.
      Lenkov suggested that Clamage chair the joint sessions this week.
      Resolution by Plum: WG21 asks Clamage to chair the joint sessions at
      this meeting.  No one objected.
4     Review and approve resolutions and issues
      Resolution 1: WG21 asks Clamage to chair the joint sessions at this
      meeting.  Approved.
5     Closing process
5.1   Select chair for next meeting
      Harbison offered to prepare the agenda for the next meeting and act as
5.2   Establish next agenda
      Jones suggested a new agenda item for the next meeting after item 5.6
      for any other new business.
5.3   Future meetings
      Deferred to the joint meeting with X3J16.
5.4   Future mailings
      Deferred to the joint meeting with X3J16.
5.5   Assign document number(s)
      Harbison asked Lajoie to assign permanent document numbers:
          MT-1 => N0730 = 95-0130
          MT-2 => SD-0
5.6   Review action items
5.7   Thanks to host
      WG21 thanked Sun Microsystems for hosting the meeting.
5.8   Adjournment
WG21 recessed at 20:55 on Sunday and reconvened in session with X3J16.  See
the corresponding WG21+X3J16 meeting minutes (N0734 = 95-0134).
Appendix A - Attendance
Name                Affiliation; (*) = Head of Delegation
Lajoie, Josee       Canada (*)
Harbison, Sam       Convener
Stroustrup, Bjarne  Courtesy
Hartinger, Roland   Germany (*)
Unruh, Erwin        Germany
Santo, Shigeru      Japan (*)
Sawatani, Yuriko    Japan
Bruck, Dag          Sweden (*)
Corfield, Sean      UK (*)
Glassborow, Francis UK
Jones, Derek        UK
Rumsby, Steve       UK
Lenkov, Dmitry      USA
Plum, Thomas        USA (*)
Koenig, Andrew      USA/Project Editor
Saks, Dan           USA/Secretary
Plauger, P. J.      WG14 Convener (ex officio)