Minutes (Not Yet Approved)
Meeting #31 of ISO/IEC JTC1/SC22/WG9

Saturday, 7 December 1996
Philadelphia, Pennsylvania, USA

Reported by James W. Moore, Convener (Acting), 703.883.7396

Call to Order: Robert Mathis, Convener

The meeting is convened at 9:00 a.m. in the Marriott Hotel by Robert Mathis, convener. He thanks SIGAda and the TriAda'96 conference committee for hosting the meeting.

Approval of Agenda

The agenda previously distributed as N321 is approved.

Administrative Arrangements: Robert Mathis, Convener

The convener reports that his retirement from the position will soon become effective, because a replacement has been identified. James W. Moore, who has acted as recording secretary for five years, has been nominated for the position. His appointment will become effective upon election by SC22. In the interim, he will act as convener. Assuming that he is elected promptly, his term will run though August 2000.

Approval of Minutes of Meeting #30

The minutes of the Montreux meeting are approved as circulated in N318.

Appointment of Meeting Chair

The Acting Convener, James Moore, is appointed as meeting chair and presides over the remainder of the meeting.

Award to Retiring Convener

The orders of the day are briefly suspended for the presentation of an award. Anthony Gargaro, representing the Special Interest Group on Ada (SIGAda) of the Association for Computing Machinery (ACM), presents the retiring convener with the Outstanding Ada Community Service Award in recognition of his accomplishments serving as convener of WG9 as well as other contributions to the Ada community. In addition, the convener is presented with a smaller award in recognition for past service as secretary of AdaTEC, the predecessor of ACM SIGAda.

In remarking upon the award, the Acting Convener notes that several of the Heads of Delegation to WG9 had participated in the awards process by providing seconds to the Mathis nomination. He recalls remarks made by the Chair of SIGAda that each of the seconds had noted additional contributions that Mathis had made to the community.

Appointment of Meeting Secretary

No individual volunteers to serve as the meeting secretary, so Moore agrees to prepare the meeting minutes and promises to resolve the situation by the next meeting. [See ACTION ITEM 31-1.]

National Body Reports and Introductions

The heads of delegation introduce themselves and other meeting participants:

In a brief report, Germany notes that 24,000 copies of an Ada CD-ROM had been distributed in Germany as an attachment to a computing magazine.

Convener's Remarks

The convener moves directly to a summary report of the SC22 plenary meeting.

Report of SC22 Plenary Meeting, Convener

[A full report of the plenary was circulated as N322.] The primary subject of discussion at the SC22 plenary was how to achieve wider public recognition of the role of ISO standards. The issue regarding copyrights was discussed extensively. ISO now requires that all working documents at all levels carry ISO copyrights. In general, participants at both the SC22 and JTC1 levels believe that this is an undesirable policy because it will hurt public access to the ISO process, hence decreasing public recognition of ISO's role. JTC1 is planning to protest the policy changes to ISO.

The convener also notes that SC22 continues to move toward greater use of electronic distribution methods. The SC22 Secretariat has announced that they are willing to accept email ballots.

Review of Old Action Items

Open action items are reviewed. Their disposition, as well as a summary of new action items, are listed in an appendix to these minutes.

Scheduling of Meeting #32

The United Kingdom states that WG9 is welcome to conduct Meeting #32 on 2 June 1997 in conjunction with the Ada-Europe conference in London. Canada proposes that the meeting be conducted in conjunction with the JTC1/SC22 plenary, 18-22 August 1997, in Ottawa.

During discussion, it is noted that broader issues of meeting frequency are relevant. The question of selecting from the two offers is broadened to one of selecting between two possible meeting schedules:

Schedule A:

Schedule B:

The discussion centers upon which of the two schedules provides a more appropriate frequency of meetings. Schedule B is adopted on a 4-2 vote of national bodies.

It is informally decided that Rapporteur Groups should be encouraged to meet in Ottawa on the proposed dates. [See ACTION ITEM 31-3.]

The Acting Convener thanks Canada for its fine proposal.

Report of ASIS Rapporteur Group: Currie Colket, Chair

Project 22.15291 (NWI ASIS)

[The ASIS Working Draft is available through both the ftp site: "" in "public/AdaIC/work-grp/asiswg/asis/v2.0". Or through the ASIS Home Page "".]

[At Meeting #30, WG9 determined that the method for approval of the ASIS Working Draft would be a 60-day letter ballot. The ASIS Working Draft had not been submitted for ballot earlier as it was not ready.]

A status is presented as follows:

The ASIS Working Draft consists of 3 Microsoft Word files:

This version is denoted ASIS Version 2.0.k, dated 7 December 1996. A Microsoft Word viewer is available in the public domain through:

Floppy disks containing the ASIS Working Draft are distributed at the meeting. ASIS is now ready for the letter ballot. Version 2.0.k will be posted to the ASIS Home Page by 10 December 1996 in word, postscript, and ASCII versions.

The ASISRG chair expresses thanks to all those who had worked so hard to produce the ASIS Working Draft. The ASIS Working Draft was developed as a volunteer effort and reflects highly on the cooperative spirit of the Ada community.

Session of the Committee of the Whole: Discussion of ASIS

The Working Group recesses into a Committee of the Whole. Dr. Strohmeier leads a discussion on issues related to ASIS. (The issues had been phrased by Dr. Strohmeier in an earlier note to WG9.):

1. Positive and negative impacts of a standard

Issue: Some people do not care to have ASIS, others do. Would the former ones be impacted negatively by making ASIS a standard? Would those who do care to have ASIS be impacted negatively by deferring ASIS standardization?


  1. ASIS is viewed as a positive interface in that as a tool to support code analysis, it was highly complementary to Ada 95 for the safety-critical, mission-critical software domains. A standardized interface could be valuable to the Ada community by encouraging the development of high-quality 3rd party tools which support a variety of code analysis requirements.
  2. A possible negative impact would be if ASIS were not mature; however, two implementations of ASIS have already demonstrated tool portability for simple examples. The WG9 discussion views the excellent cooperation and acceptance amongst the major Ada implementors (as discussed earlier) as positive.
  3. The following issues concerning the existing ASIS specification are made:

2. Timeliness

Issue: What are the pros and cons to have an ASIS standard NOW or to have a better ASIS only TOMORROW. (As all WG9 members know, no standard is ever 100% ready :-)


  1. OMG has a model whereby they distribute "standards" right away so that people can be thinking about any issues. For example, an early version (version 1.1) of one of OMG's standards was distributed (although not even close to 100% complete); this was followed relatively quickly with version 2.0, which is nearly 100% complete. It is viewed that the best way to achieve a viable ASIS interface was to press on with the ASIS standardization effort.
  2. There is a concern about standardizing the ASIS interface now and finding out that additional interfaces are needed. As mentioned, the ASIS conformance rules provide several mechanisms to extend the ASIS interface for both optional and implementor specific interfaces. ASIS can easily be extended, if necessary, by adding extensions to existing packages or by adding additional packages. An implementor may want to take advantage of this capability to deal with implementor specific capabilities and dynamic semantics.

3. ISO standard or industry standard

Issue: Should ASIS be an ISO standard, or just a de facto standard, accepted by all or almost all compiler vendors. (Notice that the ASIS NWI, formally accepted by WG9, already contains the very reasons for which ASIS needs to be an ISO standard. I mention point 3 to be complete; if everybody still agrees on this point, then let's save our time, and drop the discussion about it :-)


  1. Nearly everyone agrees that ASIS should be an ISO standard or else nobody would pay attention to it.
  2. There is a discussion as to whether ASIS should be a standard or an ISO technical report. It is noted that the NWI also stated that ASIS should be a standard-not an ISO Technical Report. A Technical Report does not have the same "status" outside of ISO that an ISO standard does. Most in the Committee support the position of moving ASIS out as a standard now.

Following the session of Committee of the Whole, the meeting of WG9 reconvenes.

The Acting Convener outlined his plan for conducting the letter ballot [See ACTION ITEM 31-5]:

  1. According to JTC1 directives, letter ballots should be scheduled for 60 days plus 14 days for mailing. This ballot will be scheduled so that it will complete by March 1. This implies that it must start immediately.
  2. The convener will send a note to the WG9 e-mail distribution list asking that each National Body choosing to participate in the ballot designate a point of contact (POC). All further correspondence regarding the ballot will be pursued with the designated POCs. [See ACTION ITEM 31-4.]
  3. The convener will email the letter ballot instructions to the designated national body POCs with a copy to the entire WG9 e-mail distribution list. Those instructions will advise the POCs that their ballots should discriminate among two types of comments: those that are merely suggestions to the ASIS RG; and those that require change so that the national body will approve the document. The instructions will quote the relevant portions of the New Work Item for the project. The instructions will require response by March 1.
  4. The Project Editor will prepare a Working Draft in accordance with the comments in the ballot and prepare a Comment Disposition Report for consideration at Meeting #32. The Comment Disposition Report should be prepared and circulated by May 1.
  5. Approval of the Comment Disposition Report at Meeting #32 will authorize the Convener to forward the Working Draft for CD registration.

Report of Ada Rapporteur Group: Erhard Ploedereder, Chair

Project 22.10.01 (IS 8652, Ada Language)

The ARG chair reports that the group met twice recently, in Montreux and in Vermont. He notes that between 1.1.96 and 31.11.96 the number of comments submitted to the ARG has grown from approximately 300 (resulting in 103 Ada Issues) to approximately 650 (resulting in 175 Ada Issues). Those 175 issues are now distributed among the steps of the ARG process as follows:

The resolutions that the ARG brings to this meeting were previously distributed in document N321. From those draft resolutions, Ada Issues 123 and 128 are deleted at the request of the ARG. From the draft resolutions, Ada Issue 12 is deleted at the request of the U. S. delegation so that it may be placed into correct editorial format. The amended draft resolutions are approved by unanimous vote of all six represented member bodies. [It should also be noted that the head of the French delegation had sent a message prior to the meeting that France had no objections.] The final text of the adopted resolutions appears in N324.

The ARG Chair notes that the ISO copyright issue is a continuing concern. The Acting Convener reiterates his determination to find a satisfactory resolution. The ARG chair suggests that after the issue is resolved, a schedule for preparation of a technical corrigenda document should be determined. [See ACTION ITEMS 31-2, 31-8 and 31-9.]

A short discussion ensues on whether technical details of AIs should discussed at WG-9 meetings. The ARG chair notes that he considers WG-9 as a final quality assurance step for the processing of AIs and that selective discussion might, on occasion, reveal considerations not considered by the ARG. In particular, he would appreciate WG-9 guidance on how conservative the AIs should be in deciding on changes to the standard. He cites, in particular, AI-141 (which deals with inappropriate locations of exception declarations in Interfaces.C child packages), where the ARG see-sawed between approval of a change in Montreux (on technical grounds) and rejection of the change in Vermont (on "political" grounds). The WG-9 chair suggests that such AIs be brought explicitly to the attention of WG-9 with due advance notice by the ARG chair. [See ACTION ITEMS 31-6 and 31-7.]

Report of Uniformity Rapporteur Group: David Emery, Chair

In accordance with JTC1 Directives Section and, the work of the URG is regarded as completed with Meeting #30.

The Acting Convener thanks the former URG Chair, David Emery, and the members of the URG for their past contributions.

Report of Annex H Rapporteur Group: Brian Wichmann, Chair

[The Chair of the HRG was unable to attend the meeting, but a copy of his report is attached as Appendix 2. A revision of their guidelines draft can be found at "".]

Steve Michell is kind enough to provide an update. He states that the next HRG meeting will occur in February 1997 in France.

Report of Numerics Rapporteur Group: Gil Myers, Chair

Project 22.10.02 (IS 11430, Elementary Functions)

Project 22.10.03 (IS 11729, Primitive Functions)

Project 22.10.04 (DIS 13813, Complex Types and Functions)

Project 22.10.05 (DIS 13814, Complex Elementary Functions)

[No representative attended the meeting. Concern was expressed regarding the unresolved status of the two DIS documents. See ACTION ITEMS 31-10 and 31-11.]

Report of Real-time Rapporteur Group: Nasser Kettani, Chair

Project 22.35 (TR 11735, Real-time Extensions)

The convener reports that the document was published 14 November 1996.

In accordance with JTC1 Directives Section and, the work of the RRG is regarded as completed with Meeting #31.

The Acting Convener thanks the RRG Chair, Nasser Kettani, and the members of the RRG for their past contributions.

Report of SQL Rapporteur Group: Andreas Koeller, Chair

Project 22.31 (IS 12227, SAMeDL)

The Chair of the SRG, Andreas Koeller, was unable to attend. In his place, Rudolf Landwehr, provided a brief report of the status of ODBC (Open Database Connectivity) and its relationship to SAMeDL. He notes that a prototype SAMeDL binding to ODBC is being prepared under funding from the German Ministry of Defense. It was suggested that National Bodies should be queried for interest in this area. [See ACTION ITEM 31-12.]

Liaisons and Other Related Work

WG4 (Cobol):

[No report.]

WG11 (Language Independent Standards):

[No report. See ACTION ITEM 31-13.]

WG14 ( C ):

The convener notes that this Working Group has expressed a desire to work closely with other language WGs to specify inter-language calling conventions. The ARA or the ARG may be able to effect liaison for us.


[See resolutions of ACTION ITEMS 30-7 and 30-8.] Anthony Gargaro reports that IEEE project P1003.21 specifies interfaces for distributed communication in real-time systems. An Ada binding is planned for this.

WG16 (LISP):

Mathis reports that IS LISP was approved as an ISO standard and that the U. S. may propose similar treatment for Scheme and Common LISP. Mathis requests that he be relieved as liaison to WG16. The Acting Convener declares him relieved and thanks him for his past contributions.

WG21 (C++):

Joyce Tokar notes that Sam Harbison has resigned as convener. Tom Plumb has replaced him. The plan is to ballot a new draft in Spring 97.

SC21/WG3 (IRDS):

[No report.]


Moore notes that he can find no interest in updating IEEE Std 990 (PDL/Ada) to the level of Ada 95. The standard will probably be administratively withdrawn in 1997.

Win 32/Ada 95 Binding:

Tokar notes that this binding has been approved by the ARA ACE effort.

Joyce Tokar was appointed as the liaison to ARA/ACE.

Bob Mathis was appointed as the liaison to the SC22 Java Study Group.

New Business

Canada proposes a resolution: "WG9 thanks Robert Mathis for his work and leadership, including the behind-the-scenes efforts we could not see, for more than ten years in support of Ada standardization." The resolution is approved by acclamation.

Final Consideration of Resolutions

The resolutions are affirmed. They can be found in document N324.


The meeting is adjourned at 2:45 p. m.

Appendix 1
Action Item Summary

Old Action Items and their Status

ACTION 30-1 [Landwehr]: Landwehr will take from the URG chair the list of existing UIs thought to be applicable to Ada 95 and prepare an informal document for discussion via email.

Landwehr reports that this will be accomplished by February 15.

ACTION 30-2 [Ploedereder]: The chair of the ARG should contact the Ada IC regarding resources to produce the HTML versions of the AIs and UIs and to place them on the Web.

Closed. The Ada IC was contacted. The item is administratively closed and replaced by 31-2.

ACTION 30-3 [All member bodies]: All member bodies should respond to the convener regarding the extent of their interest in revising the SAMeDL standard or in doing an ODBC binding.

Closed. No interest has been reported. After the report of the SRG, though, the Acting Convener is asked to again query interest in the subject. ACTION ITEM 31-12 is assigned as a result.

ACTION 30-4 [Michell for HRG]: If the HRG intends to do anything more than participate in language maintenance activity, then it must prepare a New Work Item for consideration by WG9.

Closed. Michell reports that this message was carried to the HRG. Correspondence between the Acting Convener and the HRG Chair indicates that the message was understood. (See Appendix 2.)

ACTION 30-5 [Ploedereder]: Circulate the HRG report on pragma Annotate to the ARG for comments.

Closed. This was accomplished and comments made by the ARG were sent to the HRG. As a result of this (and other activities), the HRG has decided not to propose such a pragma at this time.

ACTION 30-6 [Michell for HRG]: Formulate a resolution for consideration at the next meeting of WG9 stating the action desired by the HRG.

Closed. The HRG chair has advised the Acting Convener that a New Work Item proposal will be presented at meeting #32.

ACTION 30-7 [Convener]: Determine the status of the ISO standard for the POSIX/Ada binding, IS 14519-1:1995.

Closed. Mathis reports that the standard has been approved.

ACTION 30-8 [Member bodies]: Encourage fast-tracking of IEEE P1003.5b at the SC22 level.

Closed. JTC1 has agreed to "fast track" this standard.

ACTION 30-9 [Convener]: The convener will ascertain the status of the WG11 documents.

Closed. The retiring convener has provided the appropriate SC22 reports to the Acting Convener. This item is administratively closed and replaced with 31-13.

New Action Items

ACTION 31-1 [Moore]: Identify an individual or a national body willing to assume the duties of recording secretary.

ACTION 31-2 [Moore]: Develop a plan for processing Ada Issues that provides for public access to the Issues during a discussion period while appropriately conforming with ISO copyright rules regarding the preparation of Technical Corrigenda.

ACTION 31-3 [Rapporteur Group Chairs]: WG9 encourages its Rapporteur Groups to schedule their meetings in conjunction with the SC22 plenary in Ottawa, 18-22 August 1997. Each RG chair should correspond with Steve Michell regarding their intentions.

ACTION 31-4 [Heads of National Delegations]: Determine a point of contact for the imminent letter ballot on ASIS.

ACTION 31-5 [Moore]: Conduct the letter ballot on ASIS.

ACTION 31-6 [Moore]: For meeting #32, designate an agenda item for discussion of ARG direction and an agenda item for consideration of Ada Issues related to C interfaces.

ACTION 31-7 [Ploedereder]: Prior to meeting #32, prepare an outline of issues related to C interfaces so that delegates may prepare themselves for a discussion of these issues.

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.

ACTION 31-9 [Moore]: Forward copies of defect reports from other language working groups to the Chair of the ARG.

ACTION 31-10 [Moore]: Contact Don Sando to determine the state of DIS 13813.

ACTION 31-11 [Mathis]: Submit DIS 13814 for JTC1 ballot after ensuring that there are no dependencies upon the uncompleted DIS 13813.

ACTION 31-12 [Moore]: Again query the national bodies regarding their interest in Ada bindings to SQL and ODBC. Based upon their responses, make recommendations concerning the standardization of such bindings.

ACTION 31-13 [Moore]: Examine the report of the SC22 Secretariat to summarize the status of LIS projects from WG11.

Appendix 2
Report of the Annex H Rapporteur Group (HRG)

[Prior to the meeting, the following report was distributed by Brian Wichmann, the Chair of the HRG.]

I had planned to produce an outline of the NWI for the proposed Ada Guidelines, but I have not yet been able to do this, and hence I suspect it will be too late to be useful. I had not planned to formally present a proposal to WG9 until the June 97 meeting in London.

We had a very productive meeting in Boston in October and as a result we have a clear view of the nature of the proposed Guidelines. This is available either from the minutes of our meeting or the existing draft (which just gives a contents list).

Another topic discussed with the pragma Annotate. We concluded that it was not worthwhile continuing with this proposal.

A major reason for the delay in producing an outline NWI for WG9 is that last week I discovered an ISO group working on integrity levels with which we should clearly be coordinating. This is ISO/IEC JTC1/SC7/WG9. The convenor of this group is also convenor of IEC/TC56/WG10, but our contact has been with TC65 which is responsible for IEC 1508. I would be very grateful if any WG9(Ada) member has contacts with these groups since it could be useful to us.

Currie Colket attended the last HRG meeting which I hope implies that the HRG Guidelines will be fully compatible with ASIS, which could be of great practical importance.