ISO/IEC JTC1/SC22/WG5 (Fortran)

JTC1/SC22/WG5

Welcome to the official home of Fortran Standards

Latest News

26 November 2009
Joint meeting of WG5 and J3, Tokyo, Japan, November 16-21 A joint meeting of WG5 and J3 was held in Tokyo, Japan, November 16-21, 2008.

The main item of business was to consider the ballot comments on the draft revision ( N1723 ) of Fortran 2003 and decide on any technical changes to be made. WG5's responses to the comments are contained in N1760. The main changes were to add locks to the coarray feature, see J3 paper 08-256r2. and to alter the treatment of volatile coarrays, see J3 paper 08-297r1. A revised summary of the coarray feature is in preparation (to be N1762).

Concern was expressed over the schedule ( N1693 ) and it was decided to review this at the next meeting (May 2009).

There was not enough time to consider fully the technical content of the TR on Further Interoperability of Fortran with C, so a letter ballot of WG5 is being held, see N1763 and N1761.

New documents since 6 November 2008: N1763 Letter ballot on the technical content of N1761 (Reid)
N1761 TR 29113: "Further Interoperability of Fortran with C" (Long)
N1760 Responses to CD comments (Reid)
N1758 Resolutions of the Tokyo 2008 meeting (Muxworthy)
N1757 UK National Activity Report (Muxworthy)
N1756 Some comments on the recent WG5 papers on coarrays (Bader)
N1746 Japan National Activity Report (Takata)

6 November 2008

New documents since 31 October 2008: N1755 Request for new features from MPI Forum (Rasmussen)
N1754 Implementation Issues for Coarrays (Maclaren)
N1753 Atomic LOAD/STORE intrinsics instead of VOLATILE coarrays (Donev)
N1748 Clarifications of Coarray Memory Model (Maclaren and Donev)

31 October 2008

New documents since 14 October 2008: N1752 Draft response to N1745 (VOLATILE coarrays) (Reid)
N1751 An Analysis of NOTE 8.38 (Maclaren)
N1749 VOLATILE Coarrays Break Existing Code (Maclaren)
N1747 Coarrays in the next Fortran Standard (Reid) - supersedes N1724

14 October 2008

New documents since 2 October 2008: N1745 VOLATILE Coarrays (Maclaren for UK)
N1744 Coarrays and Memory Models (Maclaren for UK)
N1743 UK proposal for Gw and G0.d Formats (UK)

2 October 2008
Ballot processing and SC22 meeting There are problems with the ballot summary, explained in N1742. In particular, the Convener is of the opinion that the US comments are not in a form that WG5 will able to use at its meeting in November. He therefore asked Van Snyder to construct a summary. This is available as N1741 and points to J3 documents that explain the details. The problems were discussed at the SC22 meeting and a satisfactory resolution was achieved, see N1742.

New documents since 17 September 2008: N1742 Convener's report from SC22 meeting (Reid)
N1741 Comments from US Member Body on CD 1539-1 (Snyder)

17 September 2008
Results of the ballots on Fortran 2008 There were formally two ballots on Fortran 2008, both closing on 31 August. The registration ballot votes, see N1739 were 9 for approval, 5 for abstention (Denmark, France, Spain, Switzerland, Ukraine), and 1 (UK) for disapproval. The CD ballot votes, see N1740 were 8 for approval, 3 for abstention (France, Japan, Spain), 3 for approval with comments (Canada, Germany, USA) and 1 (UK) for disapproval. This means that the draft is registered with ITTF, which is a formal milestone. WG5 is required to provide a response to all the comments.

New documents since 8 September 2008: N1740 Result of SC22 CD ballot on Fortran 2008 (Seitz)
N1739 Result of SC22 registration ballot on Fortran 2008 (Seitz)

8 September 2008
Corrigendum 3 approved The country votes on Corrigendum 3, see N1738 were 11 for approval, 5 for abstention, and none for disapproval. There were no comments. The corrigendum will be forwarded to ITTF for publication.

New documents since 21 July 2008: N1738 Result of SC22 ballot on Corrigendum 3 (Seitz)
N1734 Tokyo 2008 meeting: Local Arrangements (Takata)

21 July 2008
New documents since 16 June 2008: N1737 WG5 Business Plan and Convener's Report for SC22 (Reid)

16 June 2008
New documents since 27 May 2008: N1736 Liaison report from IFIP WG 2.5 (Numerical Software) (Snyder)
N1735.pdf The new features of Fortran 2008 (Reid) - supersedes N1729
N1733 Tokyo 2008 meeting: Announcement (Reid)
N1732 Tokyo 2008 meeting: Preliminary Agenda (Reid)
David Muxworthy has provided electronic versions of the minutes and resolutions of the meetings between 1985 and 1994. The directory N1097-N1100 has been renamed as N101-N1100 and the files have been added.

27 May 2008
New documents since 19 May 2008: N1729.pdf The new features of Fortran 2008 (Reid) - supersededed by N1735
David Muxworthy has provided electronic versions of the minutes and resolutions of the meetings between 1985 and 1994. The directory N1097-N1100 has been renamed as N101-N1100 and the files have been added.

19 May 2008
The draft Corrigendum 3, N1727, and the document supporting it, N1728, have been checked by J3. N1727 was accepted and some very minor changes were suggested for N1728 to bring the edits there into line with those in N1727. The new document is N1730. This and the version of the corrigendum that does not have references to page numbers and interpretations, N1731, were sent to SC22 today.

New documents since 9 May 2008: N1731.pdf Draft Technical Corrigendum 3, as sent to SC22 (Muxworthy)
N1730 Defect reports that led to Corrigendum 3 and responses to them (Whitlock) - supersedes N1728.

9 May 2008

New documents since 10 April 2008: N1728 Defect reports that led to Corrigendum 3 and responses to them (Whitlock)
N1727.pdf Draft Technical Corrigendum 3 (Muxworthy)

10 April 2008

New documents since 20 March 2008: N1726 Result of the interpretations ballot 5, N1722 (Reid)
N1725.pdf Index of meetings (SD7) (Muxworthy) - supersedes N1606
SD1 Complete Document Register (titles of papers prior to N303 provided by David Muxworthy)

20 March 2008
Committee Draft of Fortran 2008 starts its ballot The first Committee Draft of Fortran 2008 is visible as N1723.pdf and has started its CD ballot, see ISO/IEC JTC 1/SC22 N4319. The ballot ends on 31 August. This is longer than the usual ballot period and has been chosen to allow the US comments to be considered at the August meeting of J3. Please, everyone, look at this and send your comments to your national body.

Summaries of new material in Fortran 2008 Following the revisions at the February meeting, John Reid has revised his summary of coarrays, see N1724.pdf. He is planning also to write a summary of the rest of Fortran 2008, but this is not written yet.

New documents since 7 March 2008: N1724.pdf Coarrays in the next Fortran Standard (Reid) - supersedes N1697
N1723.pdf CD revision of the Fortran Standard (Cohen) - supersedes N1718

7 March 2008
New documents since 19 February 2008: N1722 WG5 letter ballot 5 on Fortran 2003 interpretations (Reid)
N1721 Result of the interpretations ballot 4, N1711 (Reid)
N1720.pdf List of hotels in Tokyo near the meeting place (Takata)
N1717 Minutes of the London 2007 meeting (Morgan) - supersedes N1691
N1715 Minutes of the Las Vegas 2008 meeting (Morgan)

19 February 2008
Joint meeting of WG5 and J3, Las Vegas, USA, February 10-15 A joint meeting of WG5 and J3 was held in Las Vegas, USA, February 10-15 2008.

The main item of business was to review the current working draft for the next revision of Part 1 of the Fortran Standard, see N1705.pdf N1705.ps.gz

Following the abstentions at the London 2007 meeting by Canada, Japan, and UK on the content of the draft revisions, the positions of these countries was considered in the hope of reaching consensus.

The delegation of Canada, see N1712, wished to reduce the size of the revision by limiting the co-rank of co-arrays to one and removing all parallel input/output. Handling co-arrays of co-rank greater than one was seen as useful in applications while not being a big implementation burden, so this was not accepted. On the other hand, parallel input/output was seen as untested and worthy of further study.

The delegation of UK, see N1710, suggested that co-arrays be moved to Technical Report and the delegation of Japan, see N1706 suggested that co-arrays be moved to another part of the standard. Neither of these ideas was acceptable to many delegates because it reduces the emphasis on co-arrays.

On the third day of the meeting, a compromise proposed by USA was found acceptable to most delegates. It is to separate parallel programming features into a "core" set that remain in Fortram 2008 while the following features are moved into a separate Technical Report on "Enhanced Parallel Computing Facilities".

The collective intrinsic subroutines.
Teams and features that require teams.
The NOTIFY and QUERY statements.
File connected on more than one image, except for the files preconnected to the units specified by OUTPUT_UNIT and ERROR_UNIT.

The technical changes needed to implement this were agreed. The draft will be available for public comment from mid March to the end of August.

The editor suggested that hyphens be removed from the words "co-array", "co-rank", etc. to conform better with the style of the document. It was pointed out that mathematics already contains similar words, e.g. "cosine" and "cotangent". The change was accepted.

The meeting made good progress on interpretations. One set of interpretations is already being voted by WG5. Another set has just completed its J3 ballot and will be voted by WG5 soon. A new corrigendum will be sent to ISO no later than June.

19 February 2008
New documents since 6 February 2008: N1719 Timetable for public consultation (Reid)
N1718.pdf N1718.ps.gz Draft revision of the Fortran Standard (Cohen) - supersedes N1705
N1716 USA National Activity Report (Snyder)
N1714 Resolutions of the Las Vegas 2008 meeting (Muxworthy)

6 February 2008
New documents since 31 January 2008: N1713 Canada National Activity Report (Xia)
N1707 UK National Activity Report (Muxworthy)
N1706 Japan National Activity Report (Takata)

31 January 2008
New documents since 23 January 2008: N1712 Reducing the extent of the co-array feature (Xia and Reid)
N1711 WG5 letter ballot 4 on Fortran 2003 interpretations (Reid)
N1710 Moving the co-array feature into a TR (Muxworthy and Reid)
N1709.pdf Draft revision of the Fortran Standard without co-arrays (Cohen)

23 January 2008
New documents since 10 October 2007: N1708.pdf N1708.ps.gz Co-arrays in the next Fortran Standard (Reid) - supersedes N1697
N1705.pdf N1705.ps.gz Draft revision of the Fortran Standard (Cohen) - superseded by N1718
N1704 Las Vegas 2008 meeting: Local Arrangements (Xia) - supersedes N1698

10 October 2007
New documents since 21 September 2007: N1698 Las Vegas 2008 meeting: Local Arrangements (Xia) - superseded by N1704
N1700 Index of documents N1651-N1700 (Reid)
N1703 Convener's report from SC22 meeting (Reid)
SD1 Complete Document Register (Reid)

21 September 2007
New documents since 30 August 2007: N1697.pdf N1697.ps.gz Co-arrays in the next Fortran Standard (Reid) - superseded by N1708
N1699 Las Vegas 2008 meeting: Preliminary Agenda (Reid)
N1701 Las Vegas 2008 meeting: Announcement (Reid)
N1702.pdf Rationale for co-arrays in Fortran 2008 (Donev)

30 August 2007
New documents since 13 August 2007: N1691 Minutes of the London 2007 meeting (Morgan)

13 August 2007
Joint meeting of WG5 and J3, London, UK, August 6-10 A joint meeting of WG5 and J3 was held in London, 6-10 August 2007.

The main item of business was to review the current working draft for the next revision of Part 1 of the Fortran Standard, see N1678.pdf N1678.ps.gz

Prompted by the UK National Activity Report N1682 and in view of there being no full implementations of Fortran 2003 yet, the meeting considered the schedule and size of the revision.

The macros facility was removed, both to reduce the size of the revision and because there is a concern that there might be a better way to add generic programming to Fortran.

The BITS type was removed since another intrinsic type was seen as a significant addition. Most of the functionality will be provided in simpler ways.

Strong support for the retention of co-arrays was expressed in view of the dominance of multiprocessors in the current market. While there was some support for moving this feature to a Technical Report, this was not accepted since it would lead to a less precise definition and probably to a delay. Further, co-arrays were seen as the 'jewel in the crown' of the revision. A very small reduction (removal of one collective procedure) was accepted.

A new timetable was accepted, N1693, which involves a delay of 10 months to the final publication. It was decided to retain the informal name Fortran 2008 since it is already in use and since the new timetable involves the final selection of technical content in 2008.

The J3 meeting made good progress on interpretations. A new corrigendum will be prepared later this year. The convener will ask SC22 for this to be formally adopted, but it will be prepared even if this request is refused. WG5 takes the view that although Fortran 2003 is 'under revision', corrigenda are needed since this standard will be the basis for implementations for several years yet.

13 August 2007
New documents since 2 August 2007: N1685 Japan National Activity Report (Takata)
N1690 Resolutions of the London 2007 meeting
N1692 WG5 Business Plan and Convener's Report to SC22 (Reid)
N1693 Timetable the next revision (Reid)
N1694 Response to Ballot Comments - Interoperability TR (Long)
N1695 Bit Manipulation Features (Long)
N1696 Dutch National Activity Report (van Waveren)

2 August 2007
New documents since 26 July 2007: N1684 US National Activity Report (Snyder)
N1686 SC22 Other Working Group - Vulnerabilities - Liaison Report (Nagle)
N1687 ISO 31-11 special functions (Nagle)
N1688.pdf N1688.ps.gz N1688.tex Special mathematical functions in Fortran (Snyder)
N1689 Letter to the Convener re co-arrays (Wallin)

26 July 2007
New documents since 19 June 2007: N1679 Updated and Simplified Repository of Requirements (Muxworthy)
N1682 UK National Activity Report(Muxworthy)
N1683 Macros should be a TR to F2008 (Donev)

19 June 2007
New documents since 5 June 2007: N1680 London 2007 meeting: Local Arrangements (Sleightholme and Chivers) - supersedes N1676
N1681.pdf Public transport in London (Muxworthy)

5 June 2007
New documents since 18 May 2007: N1678.pdf N1678.ps.gz Draft revision of the Fortran Standard (Cohen)

18 May 2007
New documents since 6 March 2007: N1676 London 2007 meeting: Local Arrangements (Sleightholme and Chivers) - superseded by N1680
N1677.pdf N1677.ps.gz Co-arrays in the next Fortran Standard (Reid) - supersedes N1675

6 March 2007
New documents since 7 February 2007: N1674.txt N1674.pdf NWI ballot result for TR on Further Interoperability (Reid)
N1675.pdf N1675.ps.gz Co-arrays in the next Fortran Standard (Reid) - superseded by N1677

7 February 2007
New documents since 17 January 2007: N1671 London 2007 meeting: Local Arrangements (Sleightholme and Chivers) - superseded by N1676
N1672 London 2007 meeting: Preliminary Agenda (Reid)
N1673 London 2007 meeting: Announcement (Reid)

17 January 2007
Publication of Corrigendum 2 for Fortran 2003 Corrigendum 2 has been published and can be downloaded without charge from ISO. It is as WG5 sent it, apart from pagination and line-wrapping changes.

17 January 2007
Final processing of standards WG5 has recently been very dissatisfied with the final processing of standards by ITTF in Geneva. Mainly, it has been caused by poor communication - changes being made by ITTF without consulting WG5. John Reid has been actively pursuing the issue through SC22 and heard that other WGs have been having similar problems. To follow this up and raise other issues of concern to SC22, the SC22 Chairman John Hill recently meet Keith Brannon in Geneva and the following procedure (first proposed by John Reid) was agreed.

1. The SC 22 project editor sends the approved format file (currently PDF) together with any additional source files necessary to the ITTF editor who verifies the text and identifies necessary changes.

2. The ITTF editor lists the changes ITTF requires and then sends them to the SC 22 project editor.

3. The SC 22 project editor constructs a new approved format and source files and sends them to the ITTF editor.

The second and third steps are repeated until no further changes are needed.

8 January 2007
Vote on TR on further interoperability The proposal has been accepted, subject to some changes to the title and the Purpose and Justification section, see N1674.pdf

17 January 2007
New documents since 25 September 2006: N1669 Co-arrays in the next Fortran Standard (Reid) - supersedes N1642
N1670 Problems re the publication of corrigenda to the Fortran Standard (Reid)
N1674.txt N1674.pdf NWI ballot result for TR on Further Interoperability (Reid)

25 September 2006
Proposal submitted for TR on further interoperability The WG5 proposal for a Type 2 Technical Report on extended interoperability with C has been submitted to SC22. For details, see N1667.

25 September 2006
New documents since 20 May 2006: 1660 Second Draft Technical Corrigendum 2 (Muxworthy) - superseded by N1664
1661 Vote on Second Draft Technical Corrigendum 2 (Reid)
1662 WG5 Business Plan and Convener's Report to SC22 (Reid)
1663 Result of vote on Second Draft Technical Corrigendum 2 (Reid)
1664 Final Draft Technical Corrigendum 2 (Muxworthy)
1665 Technical Corrigendum 2, as sent to SC22 (Muxworthy)
1666 Defect reports that led to Corrigendum 2 and responses to them (Whitlock)
1667 New Work Item: TR on Further Interoperability (Reid)
1668 Convener's notes on the ISO/IEC JTC1/SC22 2006 Plenary (Reid)

20 May 2006
Meeting of J3, Las Vegas, Nevada, USA, May 13-17 A meeting of J3 was held in Las Vegas, 8-12 May 2006.

This was particularly important for WG5 since J3 had decided that to meet the schedule, it need to commence integration of the new features into the standard after this meeting. Therefore, it had to decide at the meeting whether the second-priority items were ready.

It was decided that BITS and intelligent macros were ready but that items J3-041/042, on extended interoperability with C, were not ready. J3 requested that these be the subject of a Type 2 Technical Report. As was the case for the Enhanced Modules TR, it should be possible to work on this in parallel with work on the new Standard without afffecting its schedule. All the other second-priority items were ready.

20 May 2006
New documents since 20 February 2006: 1654 Minutes of the Fairfax 2006 meeting
1655 Japan National Activity Report (Takata)
1656 Annexe to Repository of Requirements, N1649 (Muxworthy)
1657 WG5 letter ballot 3 on Fortran 2003 interpretations (Reid)
1658 Result of the interpretations ballot, N1657 (Reid)
1659 Draft Technical Corrigendum 2 (Muxworthy) - superseded by N1660

22 February 2006
Corrigendum 1 published Corrigendum 1 has been published and is accessible without charge from ISO. Unfortunately, it has been altered without consulting WG5 (very sloppy changing of italic font to Roman or vice-versa).

20 February 2006
Joint meeting of WG5 and J3, Fairfax, Virginia, USA, February 13-17 A joint meeting of WG5 and J3 was held in Fairfax, 13-17 February 2006.

The main item of business was to consider the Repository of Requirements N1649 and N1656 and make a firm choice of significant features for the next revision of the Fortran Standard. The most significant features were co-arrays, BITS (originally called TYPELESS), and intelligent macros (replacing parameterized modules).

Co-arrays were reaffirmed in their current form, despite a request from Bob Numrich (their inventor) to simplify them.

No firm decision was made on BITS or intelligent macros. These were left as second priority, to be done by J3 if time permits.

For full details of the chosen content, see the resolutions N1653

20 February 2006
New documents since 7 October 2005: 1648 The new features of Fortran 2003 (Reid) - supersedes N1579
1649 Repository of Requirements (Muxworthy) - supersedes N1626
1650 Index for documents N1601-N1650 1651 UK National Activity Report (Muxworthy)
1652 United States National Body 2004-2005 Activity Report (Snyder)
1653 Resolutions of the Fairfax 2006 meeting

This set of pages on the World Wide Web is produced and maintained by JTC1/SC22/WG5 in order to inform any interested parties about the development of the Fortran language. It contains a summary of the present state of standardization of Fortran and of current developments which will, in due course, lead to further extensions to the language, as well as a brief history of the standardization of the Fortran language and the names of the key individuals and organisations involved:

Overview of WG5 and the Standardization Process

WG5 Officers and Project Editors

WG5 Projects

WG5 Electronic Document Archive

WG5 Email List

Future WG5 Meetings

WG5 Projects

WG5 currently has one major project allocated to it by SC22:

Maintenance and revision of Part 1 of the Fortran Standard (base language), ISO/IEC 1539-1:2004(E), informally known as Fortran 2003.

In addition, WG5 is responsible for two other published Standards and two Technical Reports:

Part 2 of the Fortran Standard (varying length strings), ISO/IEC 1539-2 : 2000(E)

Part 3 of the Fortran Standard (conditional compilation), ISO/IEC 1539-3 : 1998

Technical Report (Type 2) on Enhanced Module Facilities, ISO/IEC TR 19767 : 2005(E)

Technical Report (Type 2) on Further Interoperability of Fortran with C, work in progress.

Future WG5 Meetings

2008 (Joint meeting with J3)	February 11-15,	Las Vegas, USA.
2008 (Joint meeting with J3)	November 16-21,	Tokyo, Japan (probably).
2009 (Joint meeting with J3)	May,	USA.