Doc No: SC22/WG21/N2682 PL22.J16/08-0192 Date: 2008-06-30 Project: JTC1.22.32 Reply to: Robert Klarer IBM Canada, Ltd. [email protected]
Jean-Paul Rigault, speaking on behalf of INRIA and AFNOR, discussed arrangements for this meeting.
| Name | Country | HOD? |
|---|---|---|
| Steve Adamczyk | USA | |
| Pete Becker | USA, Project Editor | |
| Steve Clamage | USA | |
| Gabriel Dos Reis | France | |
| Francis Glassborow | UK | Yes |
| Peter Gottschling | Germany | Yes |
| Barry Hedquist | USA | Yes |
| Howard Hinnant | USA | |
| Robert Klarer | Canada | Yes |
| P. J. Plauger | USA | |
| Tana Plauger | USA | |
| Tom Plum | USA | |
| Jean-Paul Rigault | France, Meeting host | |
| Georges Schumacher | France | Yes |
| Nick Stoughton | USA, POSIX liaison | |
| Bjarne Stroustrup | USA | |
| Herb Sutter | (Convenor) |
The agenda is document PL22.16/08-0105 = WG21/N2595.
Agenda adopted by unanimous consent.
Adamczyk, Dos Reis, and Hinnant were named to the drafting committee.
The minutes from the last meeting are recorded in document J16/08-0103 = WG21/N2593.
Approved by unanimous consent.
None.
deferred to Monday session.
deferred to Monday session.
deferred to Monday session.
deferred to Monday session.
Sutter reported that SC22 has requested that a project not be open unless the WG is ready to issue a ballot relating to that item within the next 12 months. WG21 currently has four open projects; three TRs and one standard:
Sutter noted that the Decimal TR is a binding to a revision of the IEEE-754 standard, which has long been delayed, but is expected to be ratified soon. He suggested that the paper be issued for combined registration and PDTR ballot at the San Francisco meeting of WG21 in September 2008.
On the Special Math Functions Standard, P. J. Plauger reported that the document is essentially done, and that he, Benito, and Brown all believe that the committee should vote to issue it as an FCD at the next meeting.
Sutter observed that WG21 is not with 12 months of sending out a draft of TR2 or of the Modules TR, and suggests closing those projects. New Work Items can be opened when those documents are ready for registration.
Dos Reis asked whether items that are in progress but that have no Work Item will be lost or forgotten by the committee.
Stoughton indicated the agenda must indicate that these discussions are taking place, so that interested parties have an opportunity to participate.
Plauger expand on this, noting that two separate actions are required: "we need to keep a list of work that we intend to do, and we need to announce those intentions to the world through the published agenda."
Clamage agreed to add agenda items for TR2 and the Modules TR under "Future Business" in the agenda.
deferred to Monday session.
deferred to Monday session.
Clamage review some possible changes in NCITS procedures, noting that the reason this committee originally had co-located meetings is that ISO had rules about the sizes of national delegations. If all NCITS voting members were accredited as members of the US delegation, the committee would run afoul of the ISO rules, because the delegation would exceed the maximum permissible size. The historical solution was that WG21 members would attend the meetings of PL22.16 (previously J16) as guests.
Clamage went on to explain that NCITS will institute a new rule in which a guest is permitted to attend a technical meeting only once. As well, ISO no longer has rules about the sizes of national delegations. Because of these procedural changes, the committee needs to reorganize its own meeting procedures.
Plum recalled that C started as an ANSI project in 1983. Around 1987, ISO expressed an interest in this project, and an SC22 WG was established, but essentially all work was done in J11. In 1989, at the SC22 level, other countries began making technical contributions, essentially after J11 was done. When WG21 was constituted, the current process was established so that WG21 and J16 members could cooperate.
Plum suggested that each delegation have time to outside the WG21 meeting to caucus before votes are taken.
Stoughton observed that the committee can look into the procedures that other groups have adopted to solve this problem.
Glassborow suggested a procedure in which, prior to any straw poll, the head of a delegation to calls a quick poll among his delegation.
Discussion ensued.
Glassborow reported that BSI has instructed him to declare that they want to see concepts in the standard, and they want the library to exploit concepts correctly. The UK will vote no if a standard that includes concepts is put out to a vote at any time within the next two years.
deferred to Friday session.
None.
See Saturday session.