Prepared: 3 June 1998
The calling notice and agenda for this meeting were originally circulated as N340.
The 34th meeting of ISO/IEC JTC1/SC22/WG9 (Ada) is hosted by Ada-Europe in conjunction with the 1998 Ada-Europe conference.
The meeting room will be in the Atrium Conference Center, the same place as the Ada-Europe conference. The particular room will be assigned during the conference.
Additional information, including hotel information, can be found at the Ada-Europe web site, http://www.ada-europe.org/.
The meeting will begin at 9:30 am and adjourn by 4:00 pm. A lunch break of approximately 1 hour will be taken at approximately 11:45 am. A transparency projector will be available.
The meeting will begin at 9:30 a.m.
David Emery has agreed to serve as Secretary for the meeting.
From Convener: The (currently unapproved) minutes of the St. Louis meeting are available as document N339. The resolutions approved at that meeting are available as document N338. No corrections to these documents have been proposed. The convener thanks Clyde Roby for preparing those minutes.
[See Resolution 34-P1.]
[Agenda]
From Convener: Expected attendees (and some Heads of Delegation who are unable to attend) are listed below. I have not been advised of national body delegations, so I am listing only the presumed head of delegation. Other individuals who replied are listed as WG9 officers or observers:
National Body Delegations:
WG9 Officers (if not already listed):
Those sending regrets include:
Ted Baker, Ben Brosgol, Vincent Celier, Norman Cohen, Gary Dismukes, Robert Eachus, Anthony Gargaro, Jan van Katwijk, Clyde Roby, Jean-Pierre Rosen, Alfred Strohmeier, Tucker Taft, Joyce Tokar, Tullio Vardanega.
Written reports have been received from the following delegations:
From Jean-Pierre Rosen, 6 April 1998:
Report will be provided at meeting because AFNOR will not meet until June 4.
From Erhard Ploedereder, 8 May 1998:
In late April, Ada Germany held a two day conference at the DASA Site in Bremen. With an unexpectedly large attendance of more than 75 people and very good talks, the meeting was a great success.
AK9, the German counterpart of WG9, is working on the definition of a safe subset of Ada for certain safety-critical applications.
From Kiyoshi Ishihata, Head of Delegation, 8 April 1998:
At this time, I have no new information to report. JIS standardization work has not yet been started.
From Tullio Vardanega, 5 May 1998:
The only activity in the reporting period has been the continued active support to the HRG work.
From Alfred Strohmeier, Head of Delegation, 30 March 1998:
No activity in relationship with Ada standardization to report from Switzerland.
For your information:
From John Barnes, Head of Delegation, 27 May 1998: The BSI Ada panel met recently. It allocated the review of sections of the HRG document to individuals who will respond to the HRG directly.
It also discussed the general items raised by the recent meeting of the ARG.
The UK wishes to make the following points:
[Agenda]
Clyde Roby continues his fine work in maintaining WG9's web page at http://www.open-std.org/jtc1/sc22/wg9/
I was pleased to see that SIGAda AdaLetters published an article by John Barnes on the work of the ARG.
I look forward to the May/June issue which will publish the AIs to date. Our thanks go to Erhard Ploedereder for cooperating with Phil Brashear and others in arranging for this publication.
Project Editors should be aware that a new edition of Part 3 of the ISO/IEC directives was released in 1997. This document is important because it is the "style guide" for editing JTC1 standards. Each project editor should have received a copy of the document.
The new Directive applies to all FDIS registered after September 1, 1998; however, documents written in accordance with the new edition are acceptable from now.
All participants should be aware that JTC1 is considering a full revision of their Part 1 directives, the directives that describe procedures for the progress of our documents. A ballot will be taken at the JTC1 plenary in the first week of June.
WG9 has completed its work on the two numerics draft standards, 13813 and 13814. Both documents passed DIS ballot in 1997. Disposition of Comments reports were submitted to SC22. Camera-ready, publishable drafts were sent to the ISO Secretariat in January 1998. (We await the publication of the documents.) We appreciate the dedicated work of the Project Editors, Don Sando and Jon Squires.
I am pleased that we successfully completed concurrent CD registration and FCD approval of the Ada Semantic Interface Specification. Completion of this somewhat risky maneuver has enabled us to cut at least one year off the schedule for completing this document. Success is due to the dedicated work of the ASIS leadership, Currie Colket, Clyde Roby and Steve Blake. The next step is the preparation of the draft for FDIS ballot. It is possible that an ASIS standard could be published before the end of 1999.
I am pleased the New Work Item for Guidance for the Use of Ada in High Integrity Systems was approved by JTC1. The work item has been assigned to WG9 and will be pursued by the Annex H Rapporteur Group, chaired by Brian Wichmann of the UK. The project has been assigned the number 22.15942 and the UK has nominated Brian as the Project Editor.
The US Institute for Defense Analysis (IDA) has been tasked to form a committee to study "the future course of Ada compiler certification". I have been invited to join the group. Clyde Roby, an employee of IDA, will also participate. (IDA, like MITRE, is a not-for-profit organization that is solely dedicated to advising the U.S. Federal government on complex issues.) Discussion of results to date is on the agenda under new business.
[Agenda]
From convener: The schedule for meeting #35 was already adopted during the St. Louis meeting of WG9. The schedule for meeting #36 is proposed for the first time:
[See Resolution 34-P2.]
[Agenda]
From convener: This is the "To Do" list for WG9. Some are informal action items assigned to various participants. Some are formal resolutions which are not yet implemented. Some are suspense items awaiting action by other groups.
After resolution of the ISO copyright issue, determine a schedule for preparation of a technical corrigenda to the Ada standard, ISO/IEC 8652.
Status: OPEN. A request for relief from ISO copyright policy has been made, endorsed by SC22, endorsed by JTC1, and forwarded to ISO Central Secretariat for disposition. ISO has not yet responded. The issue is "on hold" pending resolution of some more general policy issues. The issues will be addressed at the JTC1 plenary meeting at Sendai, Japan, June 2-5, 1998.
The following guidance is provided to the Ada Rapporteur Group in dealing with Ada Issues: it should be possible to write portable bindings to C programs. The Ada Rapporteur Group is requested to propose an appropriate mechanism.
Status: OPEN. From Erhard Ploedereder, 4 May 1998: "The editor of the particular AI dealing with this question has not yet completed the writeup of the ARG findings. Therefore the Action Item remains open."
WG9 forwards a proposed New Work Item, "Guidance for the Use of Ada in High Integrity Systems," to ISO/IEC JTC1/SC22 and recommends that SC22 propose its approval by JTC1. WG9 is prepared to accept the responsibility for the performance of this project.
Status: CLOSED. The New Work Item was approved.
WG9 provides the following direction to the Annex H Rapporteur Group (HRG) for the performance of any project resulting from approval of the proposed New Work Item on "Guidance for the Use of Ada in High Integrity Systems":
WG9 directs the HRG to present the resolution of this Work Item in a positive fashion which includes some abstraction of the regulatory rules; the resolution shall demonstrate:
Status: CLOSED. See Report of HRG Chair in Attachments
[Agenda]
From Robert Follett, SC22 Chairman: Microsoft Word 7 has been removed from the list of acceptable formats. It had been presumed that the format generated by Word 7 was identical with the format generated by Word 6. This turns out to be untrue in some cases. The current list of acceptable formats include:
(More information is available at: http://www.open-std.org/JTC1/IMPIT/open/j14769.doc.)
The difficulties in using Word, WordPerfect or RTF for drafts of complex documents have been extensively discussed in WG9. This suggests that, in the future, we might want to distribute documents in the form of PDF.
The note at http://www.open-std.org/jtc1/sc22/wg9/PS-to-PDF.txt describes an experimental service being provided by another WG convener in SC22. This service will convert PostScript documents to PDF. You might want to try it out.
Documents circulated only within WG9 can be in any form that we find acceptable, including PostScript.
Currently, documents submitted to ISO Central Secretariat for publication may be in a wide variety of formats, but this may change.
[Agenda]
[Agenda]
From N328 and N329,
1997 Convener's Report:
Project 22.10.02 -- IS 11430:1994 Generic Package of Elementary
Functions for Ada, Ken Dritz, editor
Project 22.10.03 -- IS 11729:1994 Generic Package of Primitive
Functions for Ada, Ken Dritz, editor
The contents of 11430 and 11729 are substantively subsumed by
the subsequent approval of the 1995 revision of 8652. These standards
will probably be withdrawn circa 1999.
From Convener's Email, 3/12/98: JTC1 directives require that all standards be reviewed for currency after five years have passed. At that time they should either be withdrawn, revised or reaffirmed as current in their existing form. In 1999, the following standards will reach the five year point:
Absent advice to the contrary, I will prepare a resolution requesting that these standards be withdrawn at the five year point because they were applicable to a superseded version of the Ada language standard.
From Ken Dritz, 1 May 1998: No activity to report.
[See Resolution 34-P8.] [Agenda]
From N328 and N329,
1997 Convener's Report:
Project 22.10.04 -- DIS 13813 Generic packages of real and
complex type declarations and basic operations for Ada (including
vector and matrix types), Don Sando and Ken Dritz, editors
Completion in 1998 is planned.
From Don Sando, Project Editor, 15 May 1998:
DIS 13813 was formally approved in the summer of 1997 with only a single dissenting vote. Final editing of the DIS prior to publication was completed in January 1998; a final draft and a disposition of comments report were submitted to ISO at that time. The final proof of DIS 13813 has been prepared by ISO, and should be received by the project editor in May 1998. Final review of this proof is scheduled to be complete on or before June 8, 1998.
From Don Sando, Project Editor, 27 May 1998:
The proof of DIS 13813, mailed by Bill Rinehuls, was waiting for me when I returned from an extended Memorial Day weekend. The French title has been correctly modified.
[Agenda]
From N328 and N329,
1997 Convener's Report:
Project 22.10.05 -- DIS 13814 Generic package of complex elementary
functions for Ada, Jon Squire and Ken Dritz, editors
Completion in 1998 is planned.
From Convener: DIS 13813 and DIS 13814 were overwhelmingly approved in balloting. Both ballots were approved by 20 votes in favor and 1 vote against. The national body voting against the two documents will reverse its vote when a straightforward change is made to them.
From Resolution 33-2: Convener is authorized to act for WG9 in approving the disposition of comments on the DIS ballot of 13814 and forwarding the draft to ISO Central Secretariat for publication.
From Convener: Disposition of comments and the publishable draft of the document were submitted in January 1998.
From Keith Brannon, ISO Central Secretariat: Publication is expected by the end of April 1998.
From Convener, 3 June 1998:
Page proofs have been sent to the Project Editor.
[Agenda]
From N328 and N329,
1997 Convener's Report:
Project 22.35 -- (Type 2) TR 11735:1996 EXTensions for real-time
Ada, Nasser Kettani, editor
The contents of 11735 are substantively subsumed by the 1995 revision
of 8652. This Technical Report will be withdrawn when usage of
the 1987 version of the Ada language has diminished.
From Nasser Kettani, Project Editor, 16 April 1998: No activity to report.
[Agenda]
From N328 and N329,
1997 Convener's Report:
Project 22.31 -- IS 12227:1995 SQL/Ada Module Description
Language (SAMeDL), Andreas Koeller, editor
WG9 has determined that the revision of 12227 is currently unnecessary;
this decision will be reconsidered periodically.
Convener's summary of status from previous discussion: None of the national bodies is interested in further work on SQL/Ada bindings beyond the current IS 12227 (SAMeDL) standard and the bindings included in the IS 9075 (SQL) standard.
[Agenda]
From N328 and N329,
1997 Convener's Report:
Project 22.15291-- WD 15291 Ada Semantic Interface Specification
(ASIS), Clyde Roby and Steve Blake, editors
From SC22 N2588, Ballot on combined CD Registration and Final CD: Due March 9, 1998. Concurrent CD Registration and FCD Approval for Ada Semantic Interface Specification (ASIS).
From SC22 N2676: Registration of FCD 15291 has been approved. WG9 is requested to prepare a Disposition of Comments Report and make a recommendation on the further processing of the FCD.
From Currie Colket, Chair of ASIS RG, 3 June 1998: Report (See Attachments)
[See Resolution 34-P9.] [Agenda]
From N328 and N329,
1997 Convener's Report:
Project 22.10.01 -- IS 8652:1995 Programming Languages: Ada,
Erhard Ploedereder and Bob Duff, editors
Planning for the revision of 8652 is commencing.
ACTION 31-8 [Moore]: After resolution of the ISO copyright issue, determine a schedule for preparation of a technical corrigenda to the Ada standard, ISO/IEC 8652.
Convener's summary of status from meeting #32: The ARG reported that most of its remaining backlog of Ada Issues will be resolved before the next meeting. Their plan is to produce an initial working draft of a Technical Corrigendum by July 1998 and to complete its approval and publication by the year 2000. The document would be a collection of issues and responses like the Corrigendum to the C language prepared in another working group of SC22. [This plan is delayed pending resolution of the previous issue - jwm]
Resolution passed at meeting #32: The following guidance is provided to the Ada Rapporteur Group in dealing with Ada Issues: it should be possible to write portable bindings to C programs. The Ada Rapporteur Group is requested to propose an appropriate mechanism.
All the AIs can be found on:
http://sw-eng.falls-church.va.us/AdaIC/standards/95com/ada-issues/arg-approved/
From Erhard Ploedereder, Chair of Ada RG, 4 May 1998: Report (See Attachments)
[See Resolutions 34-P10.] [Agenda]
Resolution from meeting #32: WG9 forwards a proposed New Work Item, "Guidance for the Use of Ada in High Integrity Systems," to ISO/IEC JTC1/SC22 and recommends that SC22 propose its approval by JTC1. WG9 is prepared to accept the responsibility for the performance of this project. [The submitted NWI is N332. Later it was placed in HTML format as document N337.]
Resolution from meeting #32: WG9 provides the following direction to the Annex H Rapporteur Group (HRG) for the performance of any project resulting from approval of the proposed New Work Item on "Guidance for the Use of Ada in High Integrity Systems":
WG9 directs the HRG to present the resolution of this Work Item in a positive fashion which includes some abstraction of the regulatory rules; the resolution shall demonstrate:
1. how Ada supports these abstractions
2. how other languages support these abstractions
3. usage paradigms for resolutions of these abstractions with a subset of Ada
From SC22 N2494: Due October 28, SC22 Letter Ballot - Approval of a Proposal for a NWI on "Guidance for the use of Ada in High Integrity Systems" - Type 3 Technical Report.
From SC22 N2683, Summary of Voting on Document JTC 1 N 5092, Proposal for a New Work Item on Guidance for the use of Ada in High Integrity Systems:
SOURCE: JTC 1 Secretariat
PROJECT: NP 15942
STATUS: This proposal has met the criteria for acceptance into the JTC 1 Programme of Work. SC 22 is instructed to begin development of this item taking into account the National Body comment received.
[The balloting report notes that the UK has nominated Brian Wichmann as the Project Editor.]
From WG9/HRG, 30 April 1998: Working Draft Version 3.4, Guidance for the use of the Ada Programming Language in High Integrity Systems. Available as N342.
From Brian Wichmann, Chair of Annex H RG, 9 April 1998: Report (See Attachments)
[See Resolution 34-P11.] [Agenda]
From WG9 N339, Minutes of Meeting #33:
An Interface Rapporteur Group (IRG) is to be created under the temporary chair of David Emery with membership of Vincent Celier, and others to be designated by National Bodies in the upcoming weeks. The IRG is directed to present recommendations at the June WG9 meeting to include the following:
From Convener, 3 June 1998: I believe that no other members of the IRG have been nominated by National Bodies. Therefore, the following materials are to be regarded as representing David Emery's recommendation, with the concurrence of Celier.
From David Emery, Chair of Interfaces RG, 1 June 1998: White Paper (See Attachments)
[See Resolution 34-P12.] [Agenda]
From Convener: The IEEE reaffirmation of the POSIX/Ada binding was successfully completed. It, along with the realtime supplement known as "1003.5b", were approved by the IEEE for publication. The supplement was incorporated into the base document and is available from IEEE as a single standard identified as "IEEE Std 1003.5b-1996 Standard for Information Technology--POSIX® Ada Language Interfaces--Part 1: Binding for System Applications Program Interface (API)--Amendment 1: Realtime Extensions."
From SC22 N2654: In the meantime, that IEEE standard (1003.5) together with a supplement thereto (IEEE Std 1003.5b-1996) is presently being balloted in JTC 1 under the Fast Track Procedure as DIS 14519, with a closing date of 18 March 1998."
From Ted Baker: There are several Ada language binding projects in the IEEE that do not have corresponding SC22 work items. These include:
[Ted Baker is unable to attend. A written report is provided below.]
From Ted Baker, Liaison to SC22/WG15, 28 April 1998:
The POSIX Ada Bindings, with Real-Time Extensions, ISO DIS 14519, passed SC22 fast-track ballot. A Resolution of Comment Report has been drafted, and is currently out for WG15 letter ballot.
The sole unresolvable negative vote was from Canada.
[Agenda]
From Bob Mathis, Convener, SC22 Java Study Group: WG9 members who have not already done so are encourage to participate or follow the work of the Java Study Group by subscribing to the mailing list as described on the Web page http://www.open-std.org/jtc1/sc22/jsg.
From Bob Mathis, Convener, SC22 Java Study Group: The JTC1 ballot on ECMAScript Fast Track (ISO/IEC DIS 16262) closes on 1998-04-09. . . . (Convener's note: ECMAScript is the standardized name for JavaScript. The standard is being pursued through the European Computer Manufacturers' Association, ECMA, and is being considered by JTC1 under the normal coordination arrangement between JTC1 and ECMA.) . . . I think the proper protocol is that the ballot closes, JTC1 decides a ballot resolution meeting is necessary, asks SC22 to convene such a meeting, SC22 appoints a convener, he schedules the meeting, and so forth. At the last Plenary, SC22 decided that I would be the Convener of this meeting if necessary.
From SC22 N 2707:
SUBJECT: Announcement of Ballot Resolution Meeting for ISO/IEC DIS 16262: Information technology - Programming languages, their environments and system software interfaces - ECMAScript: A general purpose, cross-platform programming language
Topic: DIS 16262 ECMAScript Date: Monday, June 15, 1998, 9:00am - 5:00pm
Location: Sun Microsystems, Menlo Park, California
Convener: Bob Mathis, bob@pithecanthropus.com
Hosts: Gary Robinson and Karl Matzke, Karl.Matzke@Eng.Sun.com
Project Editor: Mike Cowlishaw, mfc@uk.ibm.com
Currently Available Comments: http://www2.hursley.ibm.com/tc39/
Preliminary Discussion: sc22jsg@dkuug.dk, http://www.open-std.org/jtc1/sc22/JSG
There will be an ISO/IEC Ballot Resolution Meeting on DIS 16262 ECMAScript hosted by Sun Microsystems at Sun's campus in Menlo Park, California, USA, on Monday, June 15, 1998, at 9:00am local time. This location is south of the San Francisco airport and north of the San Jose airport. Details about the meeting room will be provided later (when an estimate of attendance can be made). There are numerous hotels in the area and no particular recommendations are being made. (The host could provide some assistance to anyone needing it.)
The Ballot has closed on DIS 16262 ECMAScript. The ITTF Table of Replies is currently only available in paper form, but will soon be converted to an appropriate electronic format for wider distribution. Some of the comments are already available at http://www2.hursley.ibm.com/tc39/. ECMA TC-39 (the originating group) has scheduled a meeting for May 15 to review the comments and draft preliminary replies in preparation for the ISO/IEC Ballot Resolution Meeting. TC-39's reactions will be posted at both sites mentioned above.
All ISO/IEC member bodies are encouraged to participate by sending delegations to this Ballot Resolution Meeting. Countries that voted NO are expected to send delegations. Countries that submitted comments are strongly encouraged to send delegations.
Delegation lists should be provided to the Convener (Bob Mathis, bob@pithecanthropus.com) and host (Karl Matzke, Karl.Matzke@Eng.Sun.com) sufficiently in advance of the meeting to make proper arrangements.
The comments currently available electronically can be viewed at http://www2.hursley.ibm.com/tc39/. That list will be updated with additional information as it becomes available. ISO/IEC discussion in advance of the meeting will be conducted on the SC22 Java Study Group mailing list: sc22jsg@dkuug.dk. Information about subscribing to that list is available at http://www.open-std.org/jtc1/sc22/JSG.
The agenda will consist essentially of discussion of the comments. A detailed agenda will be developed based on ECMA TC-39's initial responses to the comments. The detailed agenda will be posted at http://www.open-std.org/jtc1/sc22/JSG and mailed to sc22jsg@dkuug.dk. Potential delegates and interested national committees should subscribe to that list and/or check that Web site. Any other updated information about the meeting will also be sent to that list.
Requests for additional information and/or other questions should be directed to the Convener: Bob Mathis, bob@pithecanthropus.com (or +1-614-538-9232 or Bob_Mathis@compuserve.com).
[Agenda]
From Jim Moore: IEEE Std 990 on the use of Ada as a PDL was administratively withdrawn by the IEEE Standards Board. The standard, already reaffirmed once, was withdrawn because no interest was expressed in revising it to remain current with the Ada 95 language revision and with object-oriented programming technologies.
[Agenda]
At this time, no unfinished items of business are known.
[Agenda]
A discussion of possible new work on one or more standards providing procedures for testing the conformance of Ada language processors to the appropriate standards.
[Agenda]
All resolution numbers shown below are provisional (hence the "P") and subject to change following the conclusion of the meeting.
The minutes of Meeting #33 as contained in document N339 and the resolutions of that meeting as contained in document N338 are approved.
[Discussion.] [Agenda]
Meeting #36 of ISO/IEC JTC1/SC22/WG9 is scheduled as follows:
[Discussion.] [Agenda]
WG9 expresses its gratitude to Ada-Europe and to the Swedish National Body for their gracious accommodations in hosting Meeting #34.
[Agenda]
ISO/IEC JTC1/SC22/WG9 expresses its grateful appreciation to Don Sando and Jon Squire for their continuing service as Project Editors of DIS 13813 and DIS 13814.
[Agenda]
[Agenda]
ISO/IEC JTC1/SC22/WG9 expresses its grateful appreciation to Clyde Roby for serving as secretary of Meeting #33 and to David Emery for serving as secretary of Meeting #34.
ISO/IEC JTC1/SC22/WG9 expresses its grateful appreciation to Clyde Roby for maintaining the WG9 Web Page.
[Agenda]
ISO/IEC JTC1/SC22/WG9 continues the following Rapporteur Groups until the next plenary meeting and expresses its grateful appreciation to their chairs for their continuing service:
[Agenda]
ISO/IEC JTC1/SC22/WG9 recommends to SC22 that the following standards be withdrawn when they reach the end of their five-year life. Both standards are relevant to the 1987 version of the Ada language standard rather than the current version.
[Discussion] [Agenda]
The disposition of comments on the ASIS Committee Draft ballot is approved by WG9 and the Project Editor of project 15291 is directed to forward Version 2.0.Q of the draft to the SC22 secretariat for FDIS balloting.
[Discussion] [Agenda]
Resolution 34-P10a: ISO/IEC JTC1/SC22/WG9 approves the following AIs which have been previously approved by the ARG "as is":
Resolution 34-P10b: ISO/IEC JTC1/SC22/WG9 approves the following AIs which have been previously approved by the ARG "with minor editorial changes":
Resolution 34-P10c: WG9 considers it important that the ARG broaden its scope to consider language extensions suitable for inclusion in a future revision of the ISO standard.
[Discussion] [Report] [Agenda]
Resolution 34-P11a: WG9 endorses the nomination by the UK National Body of Brian Wichmann as Project Editor for 22.15942 and recommends his appointment by SC22.
Resolution 34-P11b: Working Draft 3.4 of Project 22.15942 is approved as a WG9 Working Draft. The Project Editor of project 15942 is authorized to forward WD 3.4 to the SC22 secretariat for CD registration.
Resolution 34-P11c: WG9 authorizes the Convener, in his judgment, to request simultaneous processing of any or all of the following steps in the progress of project 15942: WD ballot, CD registration, CD ballot, FCD ballot.
[Discussion] [Agenda]
WG9 authorizes the continuation of the Interfaces Rapporteur Group (IRG), chaired by David Emery, under the provisions of the following charter:
1. Name: This group shall be called the Interfaces Rapporteur Group (IRG).
2. Scope: This group is concerned with Ada interfaces to external systems, including the following:
a. Ada bindings to Application Program Interfaces (APIs)
b. Technical issues in interfacing Ada to other programming languages (in close cooperation with the ARG)
3. Approach: This group shall perform the following activities:
a. Provide liaison and review of Bindings-related activites that occur in other ISO SCs/WGs. (WG9 shall refer any bindings-related action to the IRG for review and comment prior to adopting a formal WG9 position.)
b. Investigate the feasability of a specification for the automatic generation of Ada bindings from other programming languages, with particular attention to C, C++ and Java. This specification shall specify the mapping from the source language to Ada, including packaging/program structure, use of language-defined pragmas, and identifiers. The intent is that running two different comforming Ada Binding tool over the same API will produce the same "visible interface" from the perspective of an applications program that "withs" the binding.
c. Provide informal review of proposed Ada bindings or other interface documents, when requested by a WG9 Member Body.
d. Assist WG9 and SC22, when requested, in processing Fast-Track standardization of Ada bindings and related interface standards.
4. Proposed NWIs:
a. Standard or Technical Report for automated binding techniques for C and C++
b. Standard or Technical Report for automated binding techniques for Java
[Discussion] [Agenda]
Adjournment is planned before 4:00 p.m.
[Agenda]
[Agenda]
[Agenda]
[Agenda]
Submitted by Erhard Ploedereder, Chair, Ada Rapporteur Group, 4 May 1998
The ARG met twice since the last WG9 meeting. The first meeting took place in St. Louis, November 14-16, 1997, and the second meeting in Burlington, MA, April 1-3. 14, resp. 12, ARG-members attended.
In general, the ARG work has been slowed down somewhat by the lack of funding. Nevertheless a number of AIs have been processed and completed.
The meeting in St. Louis dealt mainly with AIs. At the meeting in Burlington, the ARG also spent a half-day on general issues that might impact the future ARG work. As these issues are of some importance to the Ada community, the following excerpts from the draft meeting summary are provided as part of this report.
The first general issue was the role of the ARG in the context of Ada validation. The ARG is expressing some willingness towards taking responsibility for adjudicating the Ada validation tests and testing results as the DoD ends its funding for Ada validation, if no other alternative is found to act in the role of the AVO.
The second general issue was the question of language evolution, now that it is rather obvious that an Ada 0X will not be a project organized and financed by the US DoD. The ARG therefore would like to broaden its work by cautiously moving towards reviewing and preparing language extensions, as a step towards making this "the" process for the next revision of the language.
Beyond general approval by WG9 and ISO, acceptance of both these responsibilities will require the cooperation of the ARA or any other sponsoring agency on the nature on this responsibility. It also will require the funded support of a coordinator and of an editor, since this workload cannot be carried on a purely voluntary basis. In terms of on-going work, it will require more judicious screening of AIs according to their importance, in order to free up the needed time. The next ARG meeting, tentatively scheduled for late August or early September, could provide an opportunity to explore how the ARG might actually handle the workload for these new responsibilities.
The third general issue was the production of the Standard Corrigendum. Examples of other Standard Corrigenda (thanks to Kiyoshi) showed that there was no uniform style used. It was decided to minimize the effort in completing the Standard Corrigendum by listing only the AI summaries and (possibly shortened) questions of the approved binding interpretations, sorted by chapter. This may not be the ideal presentation, but it is the most time- and cost-effective method. Inclusion of the often lengthy discussion and recommendation sections was rejected as politically inopportune. The production of actual wording changes was considered too time-consuming to be an option.
The ARG would appreciate direction from WG9 regarding these general issues. One possible motion, suitably refined, might read: "WG9 considers it important that the ARG broaden its scope to consider language extensions suitable for inclusion in a future revision of the ISO standard."
As usual, the ARG brings to WG9 two motions for the approval of several AIs. The first motion deals with AIs that have been approved by the ARG "as is". They are now in editorial review to be finished by the WG9 meeting. The second motion deals with AIs that have been approved by the ARG "with minor editorial changes". It is doubtful whether these changes can be made and reviewed in time for the WG9 meeting. In the interest of timeliness, the motion is proposed now and will be appropriately modified by the time of the WG9 meeting. All AIs are in the usual place on the WWW server.
[See Resolutions 34-P10.]
[Agenda] [Discusssion]
Submitted by Currie Colket, Chair, ASIS Rapporteur Group, 3 June 1998
This report covers the ASIS activities since the last ISO/IEC JTC1/SC22 WG9 meeting on 14 November 1997 in St. Louis, Missouri; United States.
At the SC22 meeting from 18-22 August 1997 in Ottawa, Canada, Mr. Jim Moore, Convener of WG9, made the official request of SC22 to concurrently conduct the Committee Draft (CD) registration Ballot for ASIS and the Final CD Ballot for ASIS. This was approved unanimously by SC22.
ISO/IEC JTC 1/SC22 authorizes the Secretariat to conduct concurrent CD registration and Final CD letter ballots for project JTC 1.22.15291 - Ada Semantic Interface Specification (ASIS).
ASIS Version 2.0.P was hand delivered to Mr. Bill Rinehuls, the SC22 Secretariat, on 12 September 1997. He commenced the SC22 Balloting process for concurrent registration of ASIS as an ISO CD and the CD Final Ballot.
The Ballots closed on 9 March 1998 (4 months for balloting and 45 days for shipping and handling).
The ASIS Specification had a successful ballot. Fourteen nations voted to Approve FCD 15291. These nations included: Belgium, Canada, Czech Republic, Egypt, Finland, France, Ireland, Japan, Netherlands, Norway, Russian Federation, UK, Ukraine, and USA. Four nations Abstained and five nations did not vote. There were no ballots voting to Disapprove.
ASIS is now an officially registered Committee Draft. The ASIS Version with FCD comments resolved and approved by WG9 and SC22 will be the Draft International Standard (DIS). Only one more Ballot, the Final DIS (FDIS) Ballot, is needed to turn ASIS into an ISO Standard.
Two nations provided comments: Japan and the USA. The Japanese comment was editorial and became C#105. The USA comments were identified as editorial comments or technical issues. The Editorial Comments resulting from SC22's Ballot are numbered from C#093 through C#105. The Technical Issues are numbered from #082 through #093.
The Disposition of Comments was produced and posted to the ASIS Home Page on 1 June 1998. Each of the comments generated by the ISO balloting process has been resolved to the satisfaction of the ASISWG/ASISRG. The Committee Draft has been updated and is also available on the ASIS Home page for the review of WG9. The current document is referred to as asis2.0.Q, dated 1 June 1998. Pending approval of WG9 and SC22, it will become the Draft International Standard made available for the Final DIS Ballot.
ASIS Version 2.0.Q incorporates all known issues presented to the ASISRG by 9 March 1998. Since then several minor issues were accepted. There is one outstanding issue, dated 14 May 1998, to which there is no consensus for a solution: Issue #095, Semantic queries for multi-part declarations.
The issue addresses which declaration should be returned by semantic queries. In some cases, it is possible to return either the declaration as defined in the package body or the declaration as defined in the package specification. This is not currently clear in the ASIS specification. The ASISRG is currently looking at ways to resolve this issue. It may be resolved soon; then again, it may be an issue best addressed after ISO standardization.
The following resolution is requested at the WG9 meeting on 12 June 1998:
[See Resolution 34-P9]
Another important milestone has occurred in the life of ASIS. Back in 1992, Al Kopp and Jim Bladen, both of TeleSoft, speculated that ASIS could be used to reconstruct the source code of an entire Ada partition simply by making ASIS queries into the compilation environment. This speculation has now turned into reality. The Quality-for-ASIS: A Portable Testing Facility for ASIS paper presented at Ada-Europe'98 describes the development of an extensive testing facility for ASIS implementations. Of particular interest is the approach taken for the testing of the generic procedure Traverse_Element. The strategy is to compile the ACVC test suite, which contains all possible Ada 95 syntactic elements, and reconstruct the ACVC using only ASIS queries. The paper reported: "The implementation of Traverse_Element in ASIS-for-GNAT passed successfully this test." This is a very important achievement as it is a clear indicator of the maturity of ASIS. This test demonstrates that ASIS implementations can correctly handle every Ada 95 syntactic element, as evidenced by reconstructing the ACVC using ASIS queries.
Pending acceptance of the Disposition of Comments by WG9 and SC22, the ASIS Draft International Standard (DIS) will be subject to only one more ballot: the Final DIS (FDIS). This ballot is for 2 months and will be a thumbs-up or a thumbs-down ballot - no comments will be allowed. Assuming the standard 45 days for shipping and handling, and the 2 months of ballot, we could have an ASIS ISO Standard by November 1998. This would be very nice to have for SIGAda'98!
Please check out the ASIS Home Page. Besides having the ASIS 2.0.Q specification, there is also a new section containing ASIS 95 tutorials. Please also follow the link to the ISO Committee Draft and the link thanking the many volunteer people and organizations who helped make ASIS a reality. The ASIS Home Page is at =>
Respectfully submitted,
Clyde G. Roby, Jr., Project Co-Editor, ISO/IEC 15291 Steven Blake, Project Co-Editor, ISO/IEC 15291 William Currie Colket, Chair, ASIS Rapporteur Group (ASISRG), ISO/IEC 15291[See Resolution 34-P9.]
[Agenda] [Discussion]
Submitted by Brian Wichmann, Chair, Annex H Rapporteur Group, 9 April 1998
Response to WG9 Resolution 32-6:
'WG9 directs the HRG to present the resolution of this Work Item in a positive fashion which includes some abstraction of the regulatory rules; the resolution shall demonstrate:
1. how Ada supports these abstractions
2. how other languages support these abstractions
3. usage paradigms for resolutions of these abstractions with a subset of Ada.'
The full response to this resolution will include the latest version of the Guidelines which we plan to distribute to WG9 on 30th April, after a full review by the HRG (i.e. all those on the HRG e-mail server).
The key approach to the issue of showing that Ada satisfies the regulatory rules (and is appropriate in contexts not covered by explicit rules) is that of verification techniques. Section 3 of the Guidelines lists and analyses those techniques required for critical systems *without* assuming Ada is being used. Hence this section provides the abstractions required.
Section 4 then considers the language questions in general terms, showing specifically how Ada has been designed to facilitate the use of the verification techniques. This section makes it abundantly clear that Ada has properties which aid verification and is a sufficient basis for the reader to draw parallels with non-Ada languages. The document avoids discussing non-Ada languages in any detail for three important reasons:
Section 5 defines three classes of language features - 'included', 'allowed' and 'excluded' which reflect the ability to undertake a specific verification technique. Hence user of the Guidelines should first list the verification techniques to be used (typically derived from a sector-specific standard for safety or security), then determine from the Guidelines the set of 'allowed'/'excluded' features. This implies that we are not recommending a single Ada subset since it will depend upon the verification techniques being used.
Section 6 provides the main Guidance showing the reason for excluding a language feature and providing the additional steps that should be taken with 'allowed' features. The overall impression of the language projected by the evaluation tables is exceptionally good. This demonstrates two things:
Section 7 concludes the Guidelines by considering issues arising from the use of the Ada run-time system.
Active participants in the HRG (alphabetical), including those that have attended two of the last three meetings, or have provided extensive e-mail comments on all the draft documents circulated:
| Praful V Bhansali | USA |
| Alan Burns | UK |
| Bernard Carre' | UK |
| Dan Craigen | Canada |
| Nick Johnson | UK |
| Stephen Michell | Canada |
| Gilles Motet | France |
| George Romanski | USA |
| Mark Saaltink | Canada |
| Gareth Rowlands | UK |
| Adam Tacy | UK |
| (Phil)JP Thornley | UK |
| David Tombs | UK |
| Tullio Vardanega | Netherlands |
| Brian Wichmann | UK |
Total 15.
WG9 members can judge for themselves the progress we have made. Personally, I am delighted to present a relatively short document which nevertheless covers the area in some depth. The two last meetings in Ottawa and Nordwijk were both very productive, although the attendance was less than the previous meetings. Several people who have not attended recent meetings have been very active in e-mail comments.
[See Resolution 34-P11.]
[Discussion] [Agenda]
Submitted by David Emery, Chair, Interfaces Rapporteur Group, 1 June 1998
In my opinion, the situation for Ada bindings has changed considerably over the last few years. The advent of Ada95 and its improved specification for pragma Interface (C) and the availability of binding generation tools has improved the technical situation for the production of Ada bindings by end users. On the other hand, the loss of U.S. DoD funding and the general reduction in the vendor support for Ada binding products has reduced the market for commercial Ada binding products to nearly nothing. Furthermore, from a WG9 perspective, there has been a retreat away from formal, open systems standards in favor of COTS products and loosely defined de-facto standards. Even within ISO, the new fast-track procedures have been used to 'standardize' Java as a fast-track submission by a corporation (rather than a Member Body or other SDO.)
The goal for WG9 in the area of bindings should be the same as that for WG15, applications software portability at the source code level. Thus we'd like to have a single Ada binding for a given API such as POSIX or GKS. Experience with POSIX/Ada has shown that there is substantial value to a well-engineered standard binding for both programmer usage and applications portability. However, the costs of developing "abstract" or "thick" bindings such as POSIX/Ada have not been well-received by the user community. The belief is that the costs (both staff-hours and time-to-market) exceed the benefits. The other concern of the user community is the requirement for Ada-specific documentation; the documentation for a C API is not necessarily applicable for an Ada binding to the same API.
Additionally, the 'locus' of the bindings work for a given standard has normally not been within WG9, but rather in the ISO WG with responsibility for the underlying API. This in some cases has produced Ada bindings with questionable "style" or integration with other Ada facilities.
If the user community is willing to accept the limitations of a "direct" or "thin" binding, then the automatic generation of Ada bindings has been shown to be practical. If all such bindings were generated by a single binding generation tool, then most of the applications programmer issues would be resolved via the decisions made by the tool. However, there are several bindings generation tools available for C. I believe that there are tools for C++, and tools for Java are under development.
WG9 can adopt several approaches for Ada bindings. These approachs are not mutually exclusive; WG9 should probably adopt some variation of all of these approaches.
The latter two approaches could be used for Ada bindings to APIs that are not ISO standards (e.g. Microsoft Windows).
The other alternative for WG9 is to "abandon" the bindings business (except for those bindings produced by other ISO groups). In this alternative, some other entity, such as the ARA, would have to pick up the "mission" of Ada bindings. WG9 could continue to provide informal review and guidance, but would have no formal role in Ada bindings.
Note that in no case have I recommended that WG9 actually start a project to do an Ada binding. There are some high payoff areas where it might be possible to get the requisite ISO Member Body participation, such as CORBA and ODBC. If WG9 were interested in a New Work Item for an Ada binding, then I would strongly recommend that WG9 'delegate' the work to a Member Body (MB) and have that MB draft the binding document. WG9 participants can informally review the draft. When a good draft exists, it can be brought into ISO standardization via two routes. The route I'd recommend involves National standardization of the proposal, followed by ISO Fast-Track. (This is the approach used for POSIX/Ada). The other alternative would be for the MB to bring the standard forward for the 'normal' processing within WG9. If managed correctly, there are substantial advantages for using ISO fast-track procedures for a standard that has previously been reviewed via National Body procedures and informal review than WG9 members.
[See Resolution 34-P12.]
[Discussion] [Agenda]
End of Document