ISO/IEC JTC/SC22/WG5 N1779 Minutes of Meeting of ISO/IEC JTC1/SC22/WG5 Hosted by INCITS/PL22.3, the US Member Body for JTC1/SC22 in Las Vegas, NV, USA May 4-8, 2009 List of Participants: John Reid (JKR Associates, UK) Convenor Dan Nagle (GMU, USA) PL22.3 chair Reinhold Bader (Leibniz Supercomputing Centre, Germany) Malcolm Cohen (NAG, UK) Bob Corbett (Sun, USA) Aleks Donev (Lawrence Livermore National Laboratory, USA) Bill Long (Cray, USA) Nick Maclaren (University of Cambridge, UK) Jeanne Martin (formerly Lawrence Livermore National Laboratory, USA) David Muxworthy (BSI, UK) Craig Rasmussen (Los Alamos National Laboratory, USA) Van Snyder (Caltech/JPL, USA) Stan Whitlock (Intel, USA) Jim Xia (IBM, Canada) Note: INCITS/PL22.3 is abbreviated throughout to PL22.3 and its papers are referenced as 09-nnn. They are available from http://www.j3-fortran.org/ WG5 papers are referenced as Nnnnn. They are available from http://www.nag.co.uk/SC22WG5/ 1. Opening of the Meeting The meeting opened at 08:00 on Monday 4th May 2009. 2. Opening business 2.1 Introductory remarks from the Convenor Apologies had been received from Masayuki Takata who had been prevented from travelling to the meeting because of the widespread outbreak of swine flu. The convenor thanked PL22.3 for the new hotel and meeting facilities which were an improvement on the previous Las Vegas location. He emphasized that the top priority of the meeting was to review the current draft document and that the meeting would continue until 12:00 on the Friday. 2.2 Welcome from the Host Van Snyder welcomed participants on behalf of PL22.3. 2.3 Local arrangements In view of the relatively tight schedule for this meeting there would be no formal social event. 2.4 Appointments for this meeting The drafting committee would be Reinhold Bader, David Muxworthy (chair), Van Snyder and Jim Xia. David Muxworthy would act as secretary and John Reid as librarian. 2.5 Adoption of the Agenda (N1767) The preliminary agenda was adopted. 2.6 Approval of the Minutes of the Tokyo 2008 Meeting (N1759) The minutes of the previous meeting were approved. 3. Matters arising from the minutes Ref item 5.1: Van Snyder had contacted the Computer History Museum in Mountain View, CA about the possibility of scanning the WG5 and PL22.3 documents belonging to Jeanne Martin; scanning had not proved to be practicable. John Reid said that some older (pre-1996) documents had been put in the WG5 online archive and asked that anyone with electronic copies of relevant documents should contact him. He was also to contact Walt Brainerd to ask if he had suggestions for expanding the archive. 4. Status of Tokyo 2008 Resolutions (N1758) T5. WG5 Document Archive: See item 3 above. T7. Technical Content of Fortran 2008 CD: This would be the guiding principle for discussions at this meeting. T8. Schedule for Revision: This would be considered later in the meeting. T9. Interoperability Technical Report: The letter ballot had been held. Production of a revised draft had been given low priority and the document was still being developed. 5. Reports 5.1 SC22 Matters (Convenor) There had been no SC22 meeting since the last WG5 meeting. It was known that there were errors in Technical Corrigendum 4 which was currently being balloted by SC22. It had been established that minor changes could probably be made to the document without its undergoing a second SC22 ballot. 5.2 National Activity Reports (Heads of Delegations) Canada: Nothing significant to report. Germany: Nothing significant to report. Japan: Report is in N1780 (submitted by email) UK: Report is in N1777. US: See item 5.3. 5.3 Report from Primary Development Body (INCITS/PL22.3) The main development activity in the past six months had been to do with coarrays. 5.4 Reports from other Development Bodies (Editors/Heads) See item 4 T9 above. 5.5 Liaison Reports: INCITS/PL22.11 (C): Craig Rasmussen There had been extensive email discussion on C syntax for interoperability. MPI: Craig Rasmussen The MPI Forum was meeting every three months. Work was proceeding on Fortran bindings. Some of the features proposed in the Interoperability TR would be needed. UPC: Dan Nagle Interoperability with UPC was under discussion. There had been no developments since the last meeting. IFIP/WG2.5: Van Snyder There had been no meeting of WG2.5 since the last WG5 meeting. OpenMp: Stan Whitlock (for Matthijs van Waveren) Work had started on version 4. A working group was being formed. WG23 (Vulnerabilities): Dan Nagle WG23 was still working through the comments from the SC22 ballot on the PDTR. Work was about to start on the Fortran annex. 6. Consider the latest draft of Fortran 2008 and decide on minor changes. Three PL22.3 papers had been referred to WG5 for consideration. The outcome of discussion was as follows. - Paper 09-184 "Atomic compare-and-swap": Referred to the development body for the TR on Further Coarray Features. - Paper 09-207 "Intrinsic assignment of scalar to unallocated allocatable array": Withdrawn. - Paper 09-208r1 "Suggested new feature for F2008": There was some discussion as to whether this was a syntax change, an interpretation or a new feature, leading to straw votes: 1. Should this change be made to F2008?: yes 10, no 0, undecided 4; 2. Should this also be an interp for F2003?: 0-many-some. The paper was referred to the PL22.3 data subgroup. [PL22.3 mode from 09:20 to 17:00] Paper 09-206 "Reducing the requirement on polymorphic rank equality in intrinsic assignment" had been referred to WG5. After discussion there was a straw vote: Should this change be pursued in F2008?: 6-0-8. The paper was referred back to PL22.3. ........................................................................... Tuesday [PL22.3 mode from 08:00 to 16:35] John Reid announced that a national activity report had been received from Japan and was on the server as N1780. David Muxworthy said that it had been suggested that the UK act as proxy for Japan in any formal vote. He was discussing Japan's views further by email. ........................................................................... Wednesday John Reid asked members to look at a new draft Strategic Plan, N1781-1, ready for discussion later in the day. [PL22.3 mode from 08:02 to 16:40] There was a brief discussion on the draft strategic plan. It was decided that another corrigendum for F2003 would be needed and that the proposed timetable for the TR on coarrays should be delayed by six months. There was inconclusive discussion on the content and schedule for a future language revision. A revised plan was to be produced. ........................................................................... Thursday John Reid asked members to look at the revised draft Strategic Plan, N1781-2, for discussion later in the day. [PL22.3 mode from 08:02 to 16:55] 7. Review the schedule for Fortran 2008 (N1693) There was a discussion on N1781-2. On item 1 (Fortran 2003) it was agreed that there would be a fifth corrigendum for F2003 later in 2009 and that future corrigenda would relate to F2008. On item 2 (Fortran 2008) it was agreed that the technical content of F2008 would be frozen at 09-007r2 (the outcome from this meeting) plus any subsequent editorial corrections and clarifications and resolution of inconsistencies. This would require a delay on the schedule set out in N1693. Some members favoured a six-month delay but the general opinion was for a three-month delay. Consequently the WG5 meeting planned for February 2010 would be postponed to June. Thereafter the plan would be to hold an annual meeting in June. The proposed schedules for the two TRs (items 3 and 4) were accepted. The start of a further revision (item 5) was to be reworded with no resources to be expended in the next three years. Participants were asked to review the draft resolutions in document N1778-3 for discussion the following day. ........................................................................... Friday There was a discussion on the latest revision of the draft Strategic Plan, N1781-3. Concern was expressed that there would be no WG5 meeting during development of the Interoperability TR. After discussion the schedules of both the TR development and the WG5 meetings were amended to allow for this, so moving the proposed June 2010 WG5 meeting back to February. Since the informal names of recent Fortran standards had taken the year in which technical development was complete, it was asked whether the name should now be Fortran 2009. It was thought that technical development was essentially completed at the Tokyo meeting and so the name would remain Fortran 2008. [PL22.3 mode from 08:30 to 09:30] 8. Construct a preliminary draft of the TR on further interoperability with C. There was a discussion on the Interoperability TR, centring mainly on whether optional arguments should be retained. This led to the straw vote: Remove optional arguments: 1 - retain them: 9; undecided: 3. This direction was referred to the Interoperability TR development group. 9. Consider the Fortran defect reports (interpretations) in J3-006. Interpretations had been taken as part of the PL22.3 business during the meeting. No specific actions had been required for WG5, other than that members had been made aware that there were known errors (N1783) in Technical Corrigendum 4 (N1774 and N1775), because of incorrect edits in answers to interpretations, and that the Convenor was pursuing the matter with the SC22 Secretariat. The SC22 ballot was still open, hence member bodies could specify the errors and corrections in their votes. 10. WG5 Business and Strategic Plans 10.1 Goals for 2009-2012 This item had been covered under discussion of the draft Strategic Plan earlier in the meeting. 11. Closing Business 11.1 Future meetings It had been decided (see above) that there would be annual summer meetings from 2011 onwards. The tentative offer from Germany for 2011 was accepted, that from Canada was revised to 2012 and one from the UK was provisionally accepted for 2013. 11.2 Any other business None was raised. 12. Adoption of Resolutions (N1778) Resolutions LV1, LV2, LV7 and LV8 were approved by unanimous acclaim and LV3 and LV6 by unanimous consent. Resolution LV4 (Technical Content of Fortran 2008 FCD) was approved by 9 individual votes to 0 with 4 abstentions and by four member body votes (Germany, Japan, UK, US) to none with one abstention (Canada). Resolution LV5 (Strategic Plans for WG5) was approved by 11 individual votes to 0 with 2 abstentions and by five member body votes to none. Following email discussions with Masayuki Takata throughout the meeting, Japan's proxy vote was cast in favour of all the resolutions. 13. Adjournment John Reid thanked the local hosts (PL22.3) again for their excellent support during the week. The meeting closed at 10:10 on Friday, May 8, 2009.