SC22 is the subcommittee responsible for programming languages and environments. SC22 has 21 participating member nations and 21 observing nations. I attended as the convener of one of its working groups, WG9, the one responsible for Ada. While many subjects were considered, this report focuses on the issues related to WG9.
Attending nations included Canada, Denmark, Finland, France, Germany, Japan, Netherlands, Norway, UK, and USA. In addition, one O member, Estonia, attended. Ireland gave its proxy to Canada.
The following working groups were represented by their conveners:
WG3 |
Fortran |
WG16 |
LISP |
WG9 |
Ada |
WG17 |
PROLOG |
WG11 |
Language Independence |
WG19 |
Formal Specification Languages |
WG14 |
C |
WG20 |
Internationalization |
WG15 |
POSIX |
WG21 |
C++ |
JTC1 has approved a two-year extension of the trial period for making Committee Draft documents publicly available on the web.
ISO has agreed with a JTC1 request to give JTC1 the authority to approve certain types of documents for free availability on the World Wide Web. These document types include:
The process for submitting and approving Publicly Available Specifications (PAS) as ISO/IEC standards will be built directly into the JTC1 procedures rather than be conducted as a trial period.
JTC1 has requested comment on extending its trial period permitting normative references to documents which are not ISO or IEC standards.
We approved a letter urging extension of the policy, either in a trial period or as a part of the permanent procedures. We suggested that the administrative controls on the policy be relaxed for textbooks.
The report for WG9 Ada is given first and then followed by briefer reports for the other Working Groups.
I submitted a longer report in writing. My oral report summarized the written report as follows:
"The scope of WG9 is development of standards for Programming Language Ada. We meet twice a year. Between now and the next SC22 plenary we will meet in October 1999 in Redondo Beach, CA and in June 2000 in Potsdam, Germany."
"Our primary standard is IS 8652:1995 Programming Languages: Ada. We have requested that this standard be confirmed (ISO jargon for reaffirmation of an existing standard without revision) at its five-year review. In order to perform maintenance, we are developing a Technical Corrigenda document that we plan to complete in 2000."
The request for confirmation was subsequently approved.
"Two obsolescent standards were approved for withdrawal by the 1998 plenary: IS 11430 and 11729. Both are still listed in the ISO/IEC catalog, so ITTF has not yet implemented the withdrawal."
I was advised that ITTF action should occur after the five year point, in other words, before the end of the year 1999.
"Three standards have been published since the 1998 plenary:
"One existing standard is proposed for withdrawal by this resolution of this plenary meeting:
The request for withdrawal was subsequently approved.
"Two projects are underway:
"The following deliverables are anticipated during the next 12 months:
"We maintain close cooperation with two professional organizations, the Special Interest Group on Ada (SIG) of the Association for Computing Machinery and Ada-Europe, and one vendor association, the Ada Resource Association (ARA)."
"The marketplace for Ada is undergoing some changes. The US DoD has elected to 'privatize' its activities in the area of compiler conformity assessment. WG9 stepped up to the responsibility of assuming this role by developing a conformity assessment standard so that the authority of ISO could substitute for the authority of DoD in this vital area. With the cooperation of the SC22, JTC1 and ITTF, WG9 has succeeded in processing an appropriate standard at the ISO 'speed of light.' Publication is anticipated in October 1999 only 13 months after SC22 endorsed the NP."
"Available resources continue to decline, partially due to National Body fees. Because of these fees, Switzerland and Sweden have already ceased participation. France will probably drop out soon. Participation by the US is in danger. I have proposed a convener's session at this plenary to discuss how to maintain high quality work in the face of declining participation by experts."
The result of this discussion is described elsewhere in this report.
"We are discussing with Geneva, alternatives funding models for Ada standards. We have suggested, for example, that the Ada language standard might be freely available on the web. Vendors who offer conforming products would pay for a listing on the web page. ITTF has indicated some interest in this model."
This group continues to be largely inactive, partly due to illness of the convener.
Revision of ISO/IEC 1989:1985 continues. The CD ballot ended nearly two years ago with a large number of comments that are being resolved. Final CD ballot is expected in July 2000. Completion of the standard is planned for July 2002.
The draft revision implements an extended character set. The revision is dependent on TR 10176 for the specification of extended characters in identifiers and on draft TR 14652 for specification of attributes of characters. The revision also uses the sorting order defined by draft IS 14651.
The working group is also preparing a technical report on object-oriented class libraries for COBOL.
The Fortran working group follows a strategy of issuing technical reports describing language revisions and then incorporating the separate documents into major revisions of the base standard. At this time, these separate documents include:
The base Fortran standard, IS 1539-1:1997 is currently being revised under the informal name of Fortran 200x. Due to the complexity of some of the proposed features, the schedule has slipped by almost two years. CD ballot is now expected in October 2002 and completion in December 2004. Along the way, minor errors have been found in the technical reports. The Working Group intends to revise the TRs; they have already posted the revised text on their web site.
Optional portions of the Fortran standard are packaged as separate parts:
This group works on language-independent items related to programming languages. Their completed work includes:
Additional work is underway:
The existing standard is in three parts:
Additional work is underway:
The revision of ISO/IEC 9899:1990, Programming Language C is underway. FCD ballot has been completed and a draft forwarded for FDIS ballot.
[Due to a technical problem, the WG15 report was unavailable to me when I wrote this report. Anyone who is interested in the WG15 progress for the year should request that I send them a copy of 22n2973.]
This working group maintains a single standard, ISO/IEC 13816:1997, Programming Language ISLISP. The working group reports that they know of four implementations of the standard. Work on a revision to incorporate a C language binding will begin during the coming year.
The base standard, ISO/IEC:1995 13211-1 Prolog - Part 1: General core, is approaching its five year review. The working group intends to provide a minor revision, primarily to correct a few minor errors. FDIS ballot on draft IS 13211-2 Prolog - Part 2: Modules is currently underway and is expected to complete in December 1999.
This group is producing standards for two formal languages, VDM-SL and Z. Progress is slow.
This group is conducting a variety of projects intended to make programming language standards more appropriate for use in all countries. In the past year, they completed a second edition of TR 10176, Guidelines for the preparation of programming language standards. An amendment has been proposed to better align this standard with ISO 10646 regarding character sets. WG20 also completed TR 11017, Framework for internationalization. Work continues on several projects:
Draft IS 14651: International string ordering
Draft IS 14652: Specification method for Cultural Conventions
Draft IS 15897: Procedures for registration of cultural elements
Draft IS 15435: Internationalization API
The first listed project is their highest priority and completion is expected in 2000. The next two listed projects will provide a framework for the specification and registration of cultural conventions, easing implementation in programming languages.
In doing this work, WG20 cooperates with CEN TC304, W3C, Unicode Consortium, ISO TC46/SC2 and various groups in JTC1.
This group maintains the C++ language standard, ISO/IEC 14882, and is developed a TR on C++ Performance (real-time features).
The German delegation provided a presentation on the Pearl 90 language. Pearl is a language for writing real-time programs for embedded systems. Currently Pearl is a DIN standard, 66253-2. Further information can be found at: http://www.din.de
English language information can be found at: ftp://ftp.irt.uni-hannover.de/pub/pearl/report.pdf
A point of contact is G. Thiele, [email protected]
An Ad Hoc Group was formed to discuss the issue that I had tabled in my report. Ideas generated by group included the following:
SC22 passed a resolution requesting that SC22 O members be permitted to participate in Working Groups. Furthermore, the Ad Hoc group is continued until the next plenary meeting.
I was appointed as liaison to SC7, the subcommittee responsible for Software and Systems Engineering.
The current chair of SC22, Robert Follett, is retiring at the end of five years of service. The US has not been able to select a candidate for endorsement because of a competition between Sun and Microsoft regarding Java.
Pending an endorsement by the US of a permanent chair, the Secretariat appointed John Hill as acting chair of SC22.
11-15 Sep 2000 |
Nara, Japan |
2001 |
Austria |
2002 |
Finland |
2003 |
Belgium |