From jwagener@amoco.com Mon Nov 15 05:41:11 1993 Received: from interlock.amoco.com by dkuug.dk with SMTP id AA21037 (5.65c8/IDA-1.4.4j for ); Mon, 15 Nov 1993 18:40:07 +0100 Received: by interlock.amoco.com id AA14297 (InterLock SMTP Gateway 1.1 for sc22wg5@dkuug.dk); Mon, 15 Nov 1993 11:41:09 -0600 Received: by interlock.amoco.com (Internal Mail Agent-3); Mon, 15 Nov 1993 11:41:09 -0600 Received: by interlock.amoco.com (Internal Mail Agent-2); Mon, 15 Nov 1993 11:41:09 -0600 Received: by interlock.amoco.com (Internal Mail Agent-1); Mon, 15 Nov 1993 11:41:09 -0600 From: jwagener@amoco.com X-Openmail-Hops: 1 Date: Mon, 15 Nov 93 11:41:11 -0600 Message-Id: Subject: informal report on X3J3 127 To: sc22wg5@dkuug.dk Cc: Private_User@amoco.com X-Charset: ASCII X-Char-Esc: 29 Item Subject: Message text JLW informal report of X3J3 meeting 127, 93/11/8-12, Albuquerque ---------------------------------------------------------------- The principal objectives of this meeting were to - process the remaining interpretations - process the B9 implementation plan (93-294) - review the X3J3 requirements process for Fortran 95 - review the editorial process for Fortran 95 Meeting 126 ended with a backlog of 18 interpretations, and 38 new ones, including the reconsiderations from the previous letter ballot, were submitted by the beginning of meeting 127. A single interpretations subgroup made good progress on these 56 items; a backlog of 36 existed at the end of meeting 127 and a letter ballot will be conducted on 19 items. All of the meeting 127 B9 deliverables from document 93-294 (aka 93-204) were met, and most meeting 128 deliverables are on target. The "A" category progressed well, and the committee favored adding nested WHERE and user elemental functions to the HPF "A" topics. The crux of item B1 (object initialization) is pointer initialization, and the committee narrowly preferred a NULLIFY([type]) intrinsic function over a .NULL. constant. The deliverables for the IEEE items (B2 and C4) were pushed back a meeting, with the new schedule calling for draft proposals at meeting 128 and 009 specifications at meeting 129. At one point the committee indicated overwhelmingly (0-1-20-0 straw vote) a preference for providing IEEE support by a collateral standard rather than directly in Fortran 95 or Fortran 2000. Items B4.1 and B4.3 (minor I/O items) moved past the tutorial stage, with draft proposals scheduled for meeting 128. The B4.2 tutorial (obsolescent features) had a bit tough sledding, with the committee requesting more emphasis on rationale and criteria for deleting features and making features obsolescent. The CPU time item (C1) will be coordinated with analogous POSIX provisions, and a corresponding draft proposal is no longer scheduled for meeting 128. The tutorials for items C2 (parameterized derived types), C3 (user functions in declarations), and C5 (exception handling) all revealed complications, with consequent straw votes indicating possible recommendations for delaying these features until Fortran 2000; however, work will continue on each. The X3J3 requirements process was reaffirmed, with maintenance of document 004, a majority vote to recommend an item for Fortran 95 or Fortran 2000, and a two-thirds vote to archive an item. On this basis, two of the 004 items were archived, one recommended for Fortran 2000, and one recommended for Fortran 95. These will be submitted to WG5 by the US Fortran TAG. In a separate action, the US TAG voted to automatically accept any X3J3 requirement recommendation for forwarding, unless an objection is received from a TAG member within a week after the X3J3 meeting. On editorial matters, conversion of the Fortran 90 standard, including WG5 technical corrigendum 1, to Frame was completed (document 007). The Frame source, a postscript version, and an ascii version should be on the X3J3 ftp server soon. A number of minor corrections have been made (proposal 1 of 93-322r1), and a number of proposed improvements approved (proposal 2 of 93-322r1). Included in the vote on proposal 2 were a couple of "under consideration" items from document 008, perhaps the most significant of which is the distribution of annex C as "footnotes" throughout the standard. Jerry Wagener jwagener@amoco.com 1993 Nov 15