--- Disclaimer: The following ASCII version of the May 1991 WG15 minutes is provided for the convenience of subscribers to this mail-list, and is not definitive. The definitive version was mailed to the Convenor earlier this year for distribution in a physical mailing. Attachments to the minutes follow in a separate posting. --- Dominic Dunlop Secretary to WG15 at May, 1991 meeting ISO/IEC JTC1/SC22/WG15 - POSIX Minutes of May, 1991 meeting Date: 1991-05-14 - 17 Location: Novotel Rotterdam Brainpark and Digital Nederland bv, Rotterdam Brainpark Chair: Herman Weegenaar (Netherlands) Vice Chair: Isak Korn (European Commission) initially; later, because Mr. Korn was unable to attend for the whole of the meeting, John Hill (US) Secretary: Dominic Dunlop (UK) Attendance: Patric Dempster Canada Randall Howard Canada Arnie Powell Canada (Head of delegation) Keld Simonsen Denmark (Head of delegation) Isak Korn European Commission (Observer invited by convenor) Jean-Michel Cornu France (Head of delegation) Gerard Stein France Christian Brack Germany Ron Elliott Germany Akio Kido Japan Yashui Nakahara Japan Shigekatsu Nakao Japan Erik van der Poel Japan Shin-ichi Yamada Japan Nobuo Saito Japan (Head of delegation) Ed Keizer Netherlands (WG14 expert) Johan van Wingen Netherlands (SC2 liaison to SC22, invited by Netherlands) Willem Wakker Netherlands (WG11 liaison to WG15) Herman Weegenaar Netherlands (Head of delegation and chair) Ralph Barker US John Hill US Jim Isaak US (Convenor) Roger Martin US Pete Meier US Donn Terry US (Head of delegation) Dan Cacon UK David Cannon UK (Head of delegation) Dominic Dunlop UK Don Folland UK Martin Kirk UK Andrew Walker UK 1. Opening of meeting 1.1. Welcome from Host The meeting was convened by the chair on behalf of NNI, the host organization, at 14:00. 1.2. Introductions & Roll call of Technical Experts Those present introduced themselves. See attendance list above. 1.3. Selection of Drafting Committee A drafting committee was formed, consisting of Patric Dempster, Martin Kirk, Roger Martin and Arnie Powell. 1.4. Adoption of agenda The agenda was accepted as appearing in attachment 2. In these minutes, additional levels of sub-headings have been added for clarity. Also, where topics were discussed out of order, the discussion is minuted under the appropriate agenda item. Issues raised during the discussion of particular resolutions are minuted in 7.1 against the relevant resolution. Because Willem Wakker could not be available on Thursday 16th, it was agreed to take the report of the ad hoc profile coordination meeting out of order. Wednesday was set aside for three break-out groups to study the topics listed below. Report-backs from these groups are minuted under the appropriate agenda item: Break-out 1: WG15 Guidelines - Interpretations - Coordination (Document WG15-N75; feedback from ad hoc) - Synchronization (Documents WG15-N121, N143 #31) - Issue #2 (Document WG15-SD3) - JTC1/TSG1 (Coordination ad hoc resolutions C, D) - Meeting location (US comments) - Letter ballot procedures Break-out 2: Response to technical comments and liaison/language binding recommendations - Real time - Security - Shell Break-out 3: New projects/new work items - POSIX Guide - JTC1 documents N141, N140, N143, N147 - SC22 documents N914, N915 - Conformance test assertions - National body comments 1.5. Approval of minutes (WG15-N118) Roger Martin pointed out that on page 8, point 1 and proposed resolution, "us" and "Us" should read "US". Christian Brack and Ron Elliott noted that their names were misspelled on attachment 1; Patric Dempster that he had been present but his name had been omitted; Jean-Michel Cornu that his electronic mail address was incorrect. Jim Isaak noted that page 1 of attachment 6 (resolutions) was missing from the mailed materials, and would be supplied to those present. The minutes were approved, noting these minor points. 2. Status, Liaison and Action Item Reports 2.1. Action Items In the following, the notation "9010-#" refers to an action from the October, 1990, meeting of WG15 (Rosario); "9006-#" to an action from the June, 1990, meeting (Paris). 9010-1: Convenor: Provide a 'white paper' on FIMS to WG15 (from Brussels) (9006-7) Complete. WG15-N137 is the result. 9010-2: Netherlands: Provide information about Language compatible arithmetic standard for circulation to WG15 Netherlands Convenor. Partially complete. Document submitted, but not circulated. See action 9105-5. 9010-3: Convenor: Document Circulation for Review and Comment P1003.3 pt1, P1003.4A, P1003.6 to be circulated to WG15 for review and comment. Partially complete: P1003.3 has been circulated; others have not. See action 9105-3 for circulation of P1003.6; circulation of P1003.4A is not necessary, as resolution 120 requests that it is included as an annex to the PDAM to IS 9945-1: 1990. 9010-4: Member bodies: Review and Comment Examine documents P1003.3 pt1, P1003.4A, P1003.6 to ensure that they are correctly formatted, written in an appropriate language/form and can be progressed to CD tpo PDAM as appropriate. (9006-21) Complete, (even where documents not officially circulated). Responses were considered by break-out group 1. 9010-5: Member bodies: Review and Comment Examine P1003.5 and P1003.9 when circulated through SC22 to determine whether they are suitable for registration as CD in October 1990 (is underlying language independent model of services sufficient i.e. both adequate and consistent). (9006-22) Complete. Responses were considered by break-out group 1. 9010-6: USA: Coordination of Security-related Activities. Review the scope of IEEE/CS POSIX security-related activities with regard to the proposed program of work and terms of reference of JTC1/SC27, JTC1/SC21, JTC1/SC18 and their working groups. Identify any areas of potential overlap. Report these findings to WG15 for consideration at its first meeting in 1991. (9006-25) Complete. WG15-N159 was circulated at the meeting. 9010-7: USA & convenor: New CD 9945-2 document. Forward a revised draft of IEEE P1003.2 for registration as a CD for 9945-2.2. (9006-34) Open. The US member body judges that P1003.2 is not yet sufficiently stable for resubmission. Johan van Wingen stated that SC2 would warmly welcome circulation of parts of the current draft relevant to character set issues as soon as possible. Action 9106-21 resulted. The issue of the forwarding of a document for CD registration was moved to the issues log, WG15-SD3, as item 4. 9010-8: Convenor: request second CD ballot to 9945-2.2 from SC22 (upon receipt of a revised draft of IEEE P1003.2). (9006-35) Complete. The convenor has made the request, although it cannot be followed through until the revised draft is available. (See status of 9010-7). 9010-9: RIN & convenor: POSIX National Profiles. Seek guidance from Working Rapporteur Group on Internationalization on improvements to the practice of WG15's handling of POSIX National Profiles. (9006-37) Open pending start of activity by SC22/WG20. Becomes 9105-2. 9010-10: Convenor: Distribute a list of SC24/WG1 documents relating to windowing. Complete. WG15-N130 is the result. 9010-11: Convenor: Assign WG15 document numbers and distribute copies of all convenor correspondence sent in response to action items, to WG15. Complete. WG15-N140 is an example. 9010-12: USA: Provide IEEE 1003.2A, Draft 6 to SC22 Secretariat for registration as PDAM to 9945-2. Open. Because of problems related to the review by SC22 of the scope of work of WG15, ANSI has taken the decision not to forward. Becomes 9105-27. 9010-13: USA: Prepare a detailed schedule for the development and synchronization of language independent and C-language binding amendments to 9945-1 for IEEE 1003.1A, 1003.4, 1003.6, 1003.8. Submit to WG15 for review and comment. Open. There is no schedule, and Paul Rabin, US expert on language independent specification, was unable to attend the meeting. Becomes 9105-24. 9010-14: National bodies: Nominate candidates for position of official liaison from WG15 to WG11 (Binding Techniques). Open. No candidates have come forward. Becomes 9105-15. 9010-15: USA: Obtain materials relating to the development of LIS for GKS, PHIGS, and SQL for use in WG15's LIS work. Complete. Paul Rabin, US expert on language independent specification, has obtained and acted on this material. 9010-16: Convenor: Develop list of WG15 work items and potential expiration dates if no progress is demonstrated and distribute this list to WG15. Complete. The list appears under item 16 of N143. There are three projects at risk: 9945-3 (system administration); Fortran bindings; and Ada bindings. Break-out group 2 considered the impact of the JTC1 two year deadline, and resolution 128 resulted. 9010-17: US member body: Provide a written report on language binding issues, such as name spaces, for use by WG11 in providing a report on the nature and extent of any problems to the 1991 SC22 plenary. (Ref. SC22 AG recommendation F.) Open. Becomes 9105-25. 9010-18: Isak Korn: Provide an annotated bibliography of known CEC, SOGITS, CEN/CENELEC, ITSTC, EWOS and other European documents of interest to WG15. Open. Becomes 9105-12. 9010-19: USA: Distribute a report describing P1003.0 and its appropriateness for use as an ISO TR. Complete. N167 was distributed to at the meeting, and considered by break-out group 3. 9010-20: Canada & Japan: Notify the security Rapporteur of the name of their security rapporteur. Japan: complete; Canada: open, but in progress. Becomes 9105-1. 9010-21: WG15 & RIN: Identify the parts of 1003.2 that affect internationalization, and submit that information to the US member body. Closed. RIN judges that internationalization issues pervade 1003.2. 9010-22: USA: Request that the P1003.2 Working Group provide those portions of the draft P1003.2 document for those parts identified by the WG15 RIN (see 9010-21). Complete. P1003.2 draft 11 was provided to RIN rapporteurs at their meeting of 91-05-13. 9010-23: USA: Initiate a dialogue in to give advanced visibility with RIN on internationalization issues, especially those pertaining to networking Complete. There was some discussion as to what the deliverable from this action might be, but it was agreed that dialogue had definitely been initiated. 9010-24: Convenor: report the results of the R & C on the P1003.5 and P1003.9 documents to the WG15 member bodies prior to the next meeting. Complete. 9010-25: Convenor: Notify RIN rapporteurs of the rescheduled Spring 1991 meeting which will coincide with the WG15 meeting. (May 13 -J17). Complete. 9010-26: Convenor: Determine the degree of participation which would be expected from South Pacific countries if the WG15 meeting were held in New Zealand (April 1992). Report to WG15 prior to May, 1991. Complete. Both New Zealand and Australia would participate if a WG15 meeting were held in their region. (See 8.1.) 9010-27: Member bodies: Determine the degree of participation which would be expected from member bodies if the WG15 meeting were held in New Zealand. Complete. Participation at meeting in Australasia: Austria: no; Denmark: yes; France: yes; Germany: yes, but concerned at cost; UK: yes; US: no specific answer. This issue was also discussed in the break-out group 1, and resolution 151 resulted. 9010-28: USA: provide information to the convenor for distribution to WG15 Heads of Delegation on how to participate in the IEEE ballot process. Complete. IEEE materials distributed as WG15-N148 at this meeting. 9010-29: Convenor: Distribute information on how to participate in the IEEE ballot process to WG15 HODs. Complete. Also part of WG15-N148. 9010-30: USA: Provide WG15 with a summary list of changes made between IEEE P1003.2 draft 9 which was distributed to WG15 for registration as a CD and the next draft which will be submitted for CD registration. This summary to be provided when the P1003.2 draft is submitted to WG15. Open. The information will be embedded as editor's notes in the draft of 1003.2 when it is circulated. Becomes 9105-28. 9010-31: Convenor: Update synchronization plan (N107) based on October 1990 discussions of WG15; redistribute to national bodies as N121; and report to SC22 for SC22 endorsement. Complete. After later consideration by break-out group 1, the amendment suggested in WG15-N143 was accepted. 9010-32: Convenor: Submit WG15 request to SC22 secretariat to provide interpretation of JTC1 procedures with regard to CD registration authority (clauses 6.5.1.1 --Jpreparation, 6.5.2.2 through 6.5.2.8 --JFinalization, and 6.5.4 --JCombined Voting). Complete. The working group may only propose CD registration to SC22; SC22 must make the decision --Jsee WG15-N132. 9010-33: USA: Notify US development body of WG15's desire to receive future draft standards of limited size with subsequent multiple amendments to those standards. Complete. 9010-34: Convenor: Set up a WG15 e-mail distribution list for use in rapid, informal distribution of WG15 material. Complete. Keld Simonsen asked if there was an address through which working group members could send mail to this list. Jim Isaak replied that there was not. 9010-35: Member bodies: Notify convener of national body e-mail address for inclusion in the WG15 e-mail distribution list. Complete. No distribution point information was received from Austria, Canada, or France. 9010-36: Convenor: Distribute a draft agenda for a WG15 ad hoc Coordination Meeting (per N075) to WG15 national bodies (deadline = November 1, 1990) Complete. 9010-37: Convenor: Distribute a meeting announcement for the WG15 ad hoc Coordination Meeting (per N075) to WG15 national bodies (deadline = November 1, 1990) Complete. The meeting was held on 1991-05-07 -J08 in Amsterdam. A report appears in 4.1 and in WG15-N163. 9010-38: Member bodies: Identify, and notify convenor of, potential participants who should be invited to the WG15 ad hoc Coordination Meeting. (deadline = December 1, 1990). Complete. 9010-39: Netherlands: Notify convenor of location of next WG15 meeting in the Netherlands (deadline = December 20, 1990). Complete. 9010-40: Member bodies: Identify, and notify to the convenor, of any persons who could regularly provide status reports to WG15 on C++, C and PCTE. Complete. The persons identified are: C++: Keld Simonsen; C: Keld Simonsen, Dave Cannon; PCTE: Hugh Davies. 9010-41: Member bodies: Review N120 and provide comments to WG15 convenor by March, 1991. Complete. 9010-42: Herman Weegenaar: Coordinate WG15 OSCRL activities between now and the May 1991 meeting of WG15. Complete. See 2.5 9010-43: Convenor: As soon as possible, circulate a typed version of N120 to WG15 member bodies for review. Complete. 9010-44: Convenor: Distribute OSCRL feedback prior to the May, 1991 meeting of WG15. Complete. WG15-N136 is the only document received. 2.2. Issues list (WG15-SD3) In the following, 9006 refers to issues first logged at the June, 1990, meeting of WG15. (See also 7.3.) 1. (9006) Advise (WG20) on multi-lingual, multi-directional presentation Closed. This issue can be forwarded to SC22/WG20, now that it has been formed. See resolution 133. 2. (9006) Where should national profiles be documented? Open, but see also 7.3 3. (9006) Canadian proposed normative changes to 9945-1 Open. 2.3. Report on IEEE TCOS standards projects John Hill proposed not to repeat the Rosario exercise of going through each group, and indicated that revised versions of IEEE TCOS SD9 and SD11 would be circulated for discussion at this meeting. The US delegation proposed that this procedure should be followed at future meetings. Randall Howard proposed that a WG15 SD number should be assigned to the document. The US would not oppose this. However, the meeting did not formally allocate a number. 2.4. Liaison reports 2.4.1. PCTE Hugh Davies reported by facsimile that ECMA 149, Abstract Specification, was approved in December, 1990. The C binding should be approved in June 1991; the Ada binding in December. There are believed to be no problems with PCTE meeting POSIX conformance criteria. 2.4.2. Security Ron Elliott reported that SC22-N920 should be forwarded to the US member body for consideration in the work of 1003.7, and for comments to be passed to SC28. Resolution 150 addresses this need. 2.4.3. C language Members with links to SC22/WG14 reported that IS 9899:1990, identical to the ANSI X3J11 C standard with the addition of a few non-normative footnotes, now exists. ANSI may withdraw its standard X3-159:1989, deferring to ISO. It is likely that ANSI will retain responsibility for interpretations, as WG14 does not have the necessary resources. Keld Simonsen stated that a proposed requirement for the support of the DIS 646 IRV was under consideration, as were the use of extended character sets in identifiers, and the issue of an extended C locale. Yashui Nakahara said that the Japanese delegation is ready to send out a proposal for a multi-byte extension to IS 9899. The document was made available to WG15 at the meeting as WG15-N171. Johan van Wingen asked whether consideration had been given to the support of DIS 10646. Keld Simonsen replied that there were problems with the handling of the null character. The length of character names is also a problem; DS has proposed a set of short identifiers. This work is being handled by WG20. 2.4.4. C++ Keld Simonsen reported that WG21 will have its first meeting in Sweden in June. Internationalization is part of its terms of reference. As yet, there is nothing further to report. 2.4.5. Language independence Willem Wakker reported that WG11 work is proceeding, with heavy input on POSIX LIS from Paul Rabin. 2.4.6. Coded character sets At the invitation of the chair, Johan van Wingen reported on SC2 activity. A revision of IS 646 is at the printers: it specifies the IRV to which 9945-1:1990 refers. Voting on DIS 10646 ends on 6th June, and may result in the inclusion of controversial features, such as floating diacritical marks. WG15 delegations should ensure that their national member bodies vote in a manner they consider most suitable for use in POSIX-conforming systems. Johan van Wingen is of the opinion that floating diacritical marks, which imply variable-width characters (that is, characters having a variable number of bytes), are unworkable. There was some discussion on this issue. 2.5. OSCRL Activity John Hill reported that the OSRCL rapporteur group had examined WG15-N120, and proposed a number of corrections to typographical errors and the filling in of blank references. The OSCRL experts identified in WG15-N120 were invited to attend the May WG15 meeting, but had chosen not to come. This indicated that there is insufficient interest to merit progressing OSCRL to a type 2 TR. It was proposed that WG15, having examined the documentation, and having established that there is insufficient interest, developed a resolution stating to SC22 that the work item was complete. This resolution, 112, and the accompanying resolution 113, should be brought to the attention of national member bodies in order that they can communicate their views to SC22 at its August plenary. 2.6. TSG1 Activity Jim Isaak informed the group that the study group's final report is in JTC1 distribution. It will be voted on at the October plenary of JTC1. The report contains eleven recommendations, most of which affect the work of WG15. Additionally, annex A addresses the issues of verifiability, portability, and a summary of portability characteristics. WG15 is not currently addressing the last of these. Actions 9105-6, 9105-14 and 9105-22 deal with issues related to the review of and comment on the document by WG15 and the US member body. 2.7. Rapporteur Group report/status 2.7.1. Security Ron Elliott reported from the rapporteur group on security (RSEC). A resolution requests US member body to provide 1003.6 draft to SC22 for review and comment. A cover sheet will note that between the document sent out for IEEE ballot --Jthat is, the document sent to SC22 -- and its recirculation, the format of document will be reworked to match IEEE SD12 (also W15-N162) document structure plan, creating two separate documents while retaining technical content. In response to SC22-N891, SC27/WG1 liaison statement, the rapporteur group recommended the appointment of Kevin Murphy as WG15 liaison. Resolution 149 formalizes this, and requests SC27/WG1 to appoint a liaison to SC22/WG15. In WG15-N144 (also numbered SC22-N920), SC18/WG4 requested comments on proposals for security requirements in printing. This is not in scope for 1003.6; resolution 150 requests that it be forwarded to US member body, which should in turn pass it to the IEEE 1003.7 working group for comment. (See also 2.4.2.) 2.7.2. Conformance Testing Roger Martin walked the group through resolutions proposed to WG15 by the rapporteur group on conformance testing (RCT). 1003.3 having been published as an IEEE standard, RCT proposed that it be used as the base document for a NP in SC22. Resolution 129 resulted. The WG15 convenor should work with the IEEE project editor to ensure that the format of the document is suitable as a basis for WG15 work. It was suggested that the US member body should also be asked to prepare a report on the interaction between language independent specifications and conformance testing, but no action item resulted. (This work is already in progress in the US; it is hoped to bring results forward in Stockholm.) The rapporteur group has looked at Test Environment Toolkit (formerly Phoenix). Conformance Test System, a newer product addressing the same need, has also been examined, with the result that resolution 130 requests that the respective developers and other interested parties to attend a WG15-convened workshop with the objective of coordinating efforts. RCT minutes will henceforth be distributed as WG15 documents; no other RCT documents will be distributed, but a document list will be attached to the minutes so that those with an interest may request documents from RCT. 2.7.3. Internationalization Keld Simonsen reported that the WG15 rapporteur group on internationalization (RIN) has decided to establish contact with other groups concerned with internationalization so as to identify overlaps in programmes of work. The group had discussed a number of technical issues relating to WG15 documents, particularly the new base document which will be brought forward for 9945-2. (See issues list, 7.3, item 4). Jim Isaak suggested action 9105-18 on Keld Simonsen and Johan van Wingen to canvass the proposed POSIX definition of the word "byte" within SC2. A first draft of a document giving guidelines for the production of national profiles was discussed; work on the collection of locale information continues, as does the preparation of a questionnaire intended to identify new areas in which cultural differences exist. Two resolutions, 133 and 134, were brought forward to WG15. The group will meet in Stockholm for 2.5 days, including, if possible, Sunday. 3. Balloting Activities 3.1. 9945-1 Status Martin Kirk reported from break-out group 2. While it must be recognized that 1003.4 does not address "hard" real-time requirements, the group recommended that 1003.4 is registered as a PDAM to 9945-1, provided that work is allocated to WG15. AFNOR's comments on the document should be forwarded to 1003.4. WG15 should request that 1003.4 and 1003.4A are merged for the purposes of the work of WG15. (See resolution 120.) 3.1.1. LIS status Donn Terry reported that, following funding by the IEEE of the creation of a Language-Independent Specification corresponding to the services defined in IS 9945-1 (1003.1 LIS), work was well advanced; it was hoped to bring the document, and the corresponding C-language binding, 1003.16, to IEEE ballot early in 1992. He felt, however, that it was too early to circulate either document to WG15 or to SC22. Action 9105-29, which requires circulation of the first three chapters of 1003.1 LIS, in order that experts can review the principles underlying the specification. 3.2. 9945-2 Status There was no significant discussion of this topic. See, however, the report on action 9010-7. 3.3. (omitted) 3.4. 9945-3 Status There was little discussion of system administration issues. A review by break-out group 2 showed that it was likely that progress would be shown before the two-year project review period expired. 3.5. Language Binding Status: 1003.5, 1003.9, 1003.16 status As regards Ada & FORTRAN, break-out group 1 agreed with the US position that material will not be brought forward until LIS is complete, and new ISO standards for languages in question exist. The Fortran binding brought forward at that time should be to Fortran 90, not to the FORTRAN 77 subset thereof. Resolutions 123 to 126 address this and related matters. Following an offer by the French delegation, resolution 126 states that the 1003.1 LIS draft should be sent from IEEE to AFNOR for comment on its suitability for Ada. These considerations mean that Fortran and Ada work in WG15 may not show progress before the two-year project review deadline. Resolution 128 addresses this issue. The Modula 2 binding development proposal of WG15-N142 was welcomed; its proposers should work with IEEE to produce a thin binding, presenting results to WG15 as they become available. Work on the production of an initial dradft of 1003.16, a C binding corresponding to IS 9945-1, is well advanced in the IEEE. 3.6. CD/PDAM Registration Resolutions 138 to 147 reflect the discussion which took place with respect to document registration. 3.7. Synchronization plan The synchronization plan, WG15-N107, was accepted with small amendment suggested in N143. 4. Coordination 4.1. Report on ad hoc coordination meeting Jim Isaak reported that minutes of the ad hoc meeting had been distributed to the group as WG15-N163. Action by WG15 was requested on five topics which merited consideration by break-out group 1. Arnie Powell later reported from this break-out group, which brought resolutions 115 to 119, originating at the coordination ad hoc meeting, forward to plenary. The break-out group was of the opinion that standards documents should not contain copies of all applicable national profiles; there was need guidance from JTC1/SGFS on the the format of profile standards. (See resolution 117.) Jim Isaak commented that "they will not have any idea what [a national profile] is", so they would have to be fully briefed before being able to deliver useful guidance. Although no specific action arose, Willem Wakker is to coordinate with SGFS. (See also 6.3, and actions 9105-7 and 9105-30.) Professor Saito was concerned that the matter of profiles covered territory far broader than simply that covered by the work of WG15; hence the ad hoc group (and its proposed successor rapporteur group) should really be some way above WG15 in the JTC1 hierarchy rather than subordinate to it. Jim Isaak agreed, but said that the WG15 ad hoc group was intended simply to handle issues related to WG15; while there were, for example, issues related to SQL, WG15 is not competent to address them. John Hill and others pointed out that there was a substantive difference between the functional profiles addressed by the ad hoc group, and national profiles. Agreement on the use of terminology would be helpful. It was agreed that it was too early to propose to SC22 for forwarding to JTC1 a more formal high-level structure -- such, for example, as a new SC on profiles. Resolution 118 requests consideration of the issue of the handling of profiles in connection with the final report of JTC1/TSG1. (See also actions 9105-6, 9105-14, and 9105-22.) Resolution 119 establishes a Rapporteur Group on Coordination of Profile Activities (RGCPA) to replace the ad hoc and requests member bodies to nominate representatives. The group will meet immediately following the JTC1/SWGFS meeting in November. This meeting, in turn, will take place in conjunction with the WG15 meeting. (RGCPA may also meet in late summer if there is sufficient interest, and if meeting notice requirements can be met.) Action 9105-17 provides initial work for the group; action 9105-4 charges the convenor with the management of RGCPA until officers can be selected. 5. Requests related to work items 5.1. Request that interpretations be made a part of WG15 work Jim Isaak asked whether break-out group 1 had any guidance to him on the handling of requests for interpretations, pointing out that there was a synchronization problem between WG15 and the IEEE, and that it is by no means clear whether interpretations are within the scope of work of WG15. John Hill commented that this issue is to be discussed at the next SC22 meeting in September, and that WG15 should perhaps modify its synchronization plan to recognize the effect of interpretations on the work of WG15. Donn Terry suggested that, pending guidance from SC22 (and perhaps, should the group ask Jim Isaak to seek it, from ITTF), the group needed an interim mechanism for dealing with any request; he suggested that requests should be forwarded to IEEE. Resolutions 154 and 155 address these issues. The group also agreed that it would be useful for it to see interpretations of IEEE Std. 1003.1:1990. Action 9105-26 resulted. 5.2. Further action on P1003.0 Guide The feeling of the group was that the POSIX Guide work of IEEE P1003.0 should be brought forward as an NP for a TR. Resolutions 135 and 136 resulted. 5.3. JTC1 AG required NPs John Hill described break-out group 3's findings on the NP/NWI work in SC22 and JTC1. The bottom line, embodied in resolutions, is that WG15 should recommend that all NWIs are forwarded (with some editorial changes) to JTC1 by SC22 for approval where they have already been seen by SC22; or to SC22 for approval by this group where they have not. The net effect will be that WG15 will continue within the scope of work understood by its members, but with formal approval for that scope of work from JTC1. Resolutions 138 to 147 recommend the approval of ten NPs to JTC1. 5.4. Requests related to other division of work/NPs Break-out group 3 had reached no conclusion about the means by which conformance test could or should be included in the scope of work of WG15; debate in plenary resulted in resolutions 130 and 131. 5.5. Modula 2 binding The proposal for the development of a Modula 2 binding to POSIX services (WG15-N142) was welcomed, although, given WG15's disapproval of the "thick" binding approach of 1003.5, the proposers should be discouraged from using that document as a basis for their work, although they should cooperate with those working on language bindings within the IEEE. Resolution 127 makes these and other points. 6. New business 6.1. Bringing forward of LIS material WW expressed concern about the timing and mechanism of the replacement of C-based material with language-independent material, the action on the US member body to bring forward a timetable notwithstanding. Jim Isaak added that there were clearly problems for synchronization: for example, will C-based real-time (1003.4) move more quickly or more slowly than 1003.1 LIS? The group cannot know. Donn Terry wondered whether the interpretation of guidance from SC2 should be "C now, required to be language independent later", or "C now, permitted to be language independent later". He preferred the latter. The group agreed, and all NP resolutions (138 -J147) reflect this decision. 6.2. Review of past resolutions It was agreed that the review of resolutions should be an agenda item at future meetings, and that it would be the responsibility of the vice chair to provide a report on the topic to the meeting. Action 9105-11 resulted. 6.3. SGFS participation The SGFS chair will be informed by the convenor that Andrew Walker, Willem Wakker, Ron Elliott, and Walter Stahlecker will represent WG15 interests at the forthcoming SGFS meeting in Berlin. 7. Review/approval of resolutions, action items and issues list Ron Elliott commented that he had found the division of work among small groups and the subsequent report-backs unhelpful; he suspected that information had been lost. Donn Terry and Herman Weegenaar proposed that this comment be revisited after discussion of the resolutions. If it transpired that every issue was re-debated, guidance should be given to future chairs that non-technical break-outs should be avoided. Resolution 163 was subsequently formulated to address this issue. 7.1. Resolutions Approved resolutions appear as attachment 3 to these minutes. In the following, debate is minuted only where it resulted in substantive change to the resolution originally proposed, or raised issues not addressed by the final resolution. Unlisted resolutions were accepted without substantive change. 114: Letter ballots This resolution arose from a comment on the conduct of the recent letter ballot from the US member body. Keld Simonsen queried whether the group was allowed to make its own rules on balloting. Jim Isaak replied that it was, provided that it was consistent with JTC1 directives. Since there are no directives on letter ballots, we can do as we wish. As an aside, Jim Isaak asked when the 60 days was considered to start. John Hill said that it was from the time the document reached the hands of all member bodies by physical mail (not all bodies having electronic mail addresses). To avoid having to use physical mail as the only means of communication, it was suggested that all member bodies should designate a means of communication and a contact point. (See action 9105-13.) 115: Participation in profiling activities Professor Saito requested clarification of who was being encouraged to participate. Jim Isaak replied that this would vary between member bodies, giving examples of the different bodies which might be expected to participate in different countries. 117: Request to SGFS regarding Open System Environment (OSE) Profiles Andrew Walker queried the removal of wording from the profile coordination ad hoc's resolution on the identification of interested parties, and the invitation of these parties to participate in SGFS activities. It was agreed that the resolution should address this point. Donn Terry also queried removal of a suggested date and location (adjacent to Stockholm). Jim Isaak stated that this might have appeared to preclude earlier activity on the topic by SGFS. Professor Saito suggested that the resolution should specifically reference 1003.0 as a suitable base document for the work of SGFS. Jim Isaak proposed that the US member body should work in any forum set up by SGFS to harmonize 1003.0. Donn Terry wished to avoid a laundry list of documents in the resolution. Action 9105-30 was formulated to take care of this issue. 118: Regarding JTC1 organization and the JTC1 TSG1 report Donn Terry said original resolution from the profile coordination ad hoc consciously did not refer to SC22, as SC22 is not empowered to do what WG15 asks. The SC22 chair can bring such matters to JTC1's attention at its June meeting, however. The resolution was accepted with this amendment. 120: Registration of P1003.4 as a PDAM There was discussion of whether a draft of 1003.4A should be provided at the same time as draft 11 of 1003.4. Should these documents simply be "glued together", or should submission be delayed pending a proper merge, as requested in resolution 121? Willem Wakker considered that to submit a glued submission would be to ask for trouble. Herman Weegenaar said that the Netherlands, for one, had reviewed only 1003.4, and so would have difficulty in formulating a voting position on a glued or even a merged document at this stage. Roger Martin pointed out that the IEEE (or ANSI) had not yet made a decision on whether it wished to forward 1003.4A to WG15 anyway. Jim Isaak countered by saying that SC22 N838 (1003.4 draft 9) had included 1003.4A information as a non-normative annex. Martin Kirk proposed that this practice be continued for the PDAM registration. The group concurred, and the resolution reflects this decision. 125: Review of LIS What, asked Arnie Powell, is a "wide review"? How does this differ from a plain review? John Hill explained: it would include US language standards developers, SC22 working groups, SC22 itselfI Willem Wakker suggested that really such work was in the domain of WG11 rather than of WG15. A wide review of the content of 1003.1 LIS, on the other hand, was a suitable topic for WG15 or, as its surrogate, the US member body. This issue, and variations thereon, was discussed at length, resulting in clarification to the intent of the resolution, which was then accepted as amended. Action 9105-32 was formulated to address the issue of WG11 review. 126: Ada bindings and LIS John Hill suggested that in addition to sending P1003.1 LIS to AFNOR, it should be sent to all WG15 member bodies. The point, said Randall Howard, was that AFNOR Ada experts had volunteered to do the work --Jwork which, it appeared, could not quickly be done in the US or elsewhere. The group agreed that it would be onerous to request that all member bodies review and comment on the whole P1003.1 LIS document; instead, the first three chapters should be circulated. Action 9105-29 resulted. 129: Proposal to adopt P1003.3-1991 as a standard Dominic Dunlop noted that, in the resolution as originally proposed, no assignment for the NP, if approved, was suggested by the resolution. Wording was added suggesting to SC22 that it either be assigned to WG15, or, if considered appropriate by SC22 member bodies, be processed through the fast track procedure. Further consideration was deferred, pending further consideration by delegations and redrafting. Following discussion of the merits and demerits of the fast-track process, the redrafted resolution was accepted. 132: Document structure --Jtest methods Action 9105-19 requests a report back from the technical editor at next WG15 meeting. 135: Request for draft of P1003.0 Roger Martin queried the result of Rosario resolution 99, which asked for SC22 consideration of the matter of including 1003.0 in WG15's scope of work. Resolution 136 addresses this issue. In general, commented Ron Elliott, we should review past resolutions at each of our meetings. See discussion under 7.2 and action 9105-11. 136: Forwarding NP for P1003.0 to SC22 This resolution refers to WG15-N138, which was discussed in a break-out group, but not reviewed in plenary session. Jim Isaak reported that, in the letter ballot, four countries had approved: Canada, Germany, Japan and the UK; none had disapproved or abstained; and four had not responded. Of the latter, Denmark indicated that it wished to continue to make no response; France, Netherlands and the US that they had intended to affirm. This made the final position: approve: 7; disapprove: 0; abstain: 0; no response: 1. The resolution was accepted with editorial amendments. 137: RPC This resolution alludes to the work of SC21/WG6. The matter of liaison on RPC issues was discussed. John Hill stated that the US would be willing to establish an internal liaison to X3T5.5, or to work with somebody else on WG15 acting as a liaison. The US would not provide a liaison at the the WG15 level but thought that this would be the better solution for WG15. In the absence of a volunteer, a proposed resolution relating to the establishment of a liaison was struck. 146: Recommendation on approval of NP for security (changes to the system interface) Ron Elliott disliked the requirement in the resolution as originally proposed for a LIS, as it would considerably delay IEEE balloting on the standard and lose IEEE/WG15 synchronization; Donn Terry added that direction from SC22 and from this group requested C language documents first, with LIS to follow. Since the wording resulted from a comment on an affirmative ballot, and the comment will not be lost, Dave Cannon proposed that the requirement be dropped in this case. The change was made and the resolution accepted. 139: Recommendation on approval of NP for transparent file access It was argued that a requirement in the resolution as originally drafted for LIS would introduce a delay, although this would be unlikely to be as great as that which would result from placing the requirement on 1003.6 (resolution 146). The 1003.8 document "touches" many areas of POSIX, so the requirement would also result in a loss of synchronization. After canvassing of national positions by the chair, it was decided that the requirement should be dropped. The resolution was accepted as amended. There was concern that the group still has not seen a plan for the bringing forward of LIS'. (See action item 9105-24.) 140: Recommendation on approval of NP for protocol independent interface (PII) Note that it is proposed that this material will be brought forward as a LIS, and hence that it cannot be an addendum to 9945-1:1990. 145: Recommendation on approval of NP for shell and utility extensions Jim Isaak observed that the wording ("UNIX") of this NP exactly echoes that of the original WG15 work item, but that "POSIX" was also acceptable. 153: Forwarding comments in WG15 N146 (Ada binding) Ron Elliott said that this really should be an action, but the resolution was accepted nevertheless. 157: In regard to possible assignment of WG15 to a new SC This resolution replaces an alternative proposal which would have welcomed the assignment of WG15 to a new SC. At the request of Donn Terry, the original motion was proposed by Professor Saito for the Japanese member body. The proposal was not to stop POSIX activity, but to perform a "directory copy" action and so better provide for current and future growth. Donn Terry countered, extending the analogy to suggest that it would be like a copy to another file system, losing important links to language, language independence, and (added Dominic Dunlop) internationalization activity in SC22. Arnie Powell tentatively agreed, and felt he had insufficient information to recommend in favour of as a large a structural change as that proposed. For the UK, Dave Cannon argued that the current pragmatic approach the the standardization of existing UNIX interfaces was working well, and should not be changed. Speaking as convenor, Jim Isaak discussed concerns about obtaining the correct international responsiveness, participation and resources for POSIX work. Although WG15 seems to work well, there have been serious problems on these fronts elsewhere in SC22. He felt that these concerns merited serious consideration, and should not be obscured by the proposal for a new SC. For the Netherlands, Herman Weegenaar stated that he had a concern about functional profiling work's current lack of a home. The initially proposed resolution did not address this issue. He had no objection to moving existing POSIX work on an open systems environment component to a new SC, but such a component was just one part of any functional profile. Professor Saito recognized the distinction, but thought it wrong to preempt the decisions of SGFS and JTC1 by proposing that profiling work be included in the proposed SC. Others agreed, pointing out that other profiling work concerned standards developed in SC21 (OSI) and elsewhere, and would probably be inappropriate for the proposed SC. As chair, Herman Weegenaar summarized discussion of the proposed support of reorganization: 1) He heard support from nowhere but Japan on the proposal for the establishment a new SC. (Denmark stated no position.); 2) That the issue of profile work was open; and 3) (from Jim Isaak) That it was not clear where new component work, as identified, should be carried out. The UK proposed a replacement resolution 157, addressing the first of these issues only. After extended debate on how procedurally to handle this, the new resolution was accepted with slight amendments. 158: Management of WG15 meetings This resolution arose from the lack of advance notice of the formation of break-out groups for the discussion of substantive non-technical issues at the current meeting --Jalthough, as the chair explained, this had been a result of the lack of a room large enough for a plenary session on 1991-05-15. Professor Saito proposed the resolution. Ron Elliott, as head of a small delegation, welcomed the proposal for advance notice of small group formation. Dave Cannon strongly agreed. There was considerable discussion of issues relating to requirements for advance notice, but the existing wording was deemed adequate. Yashui Nakahara emphasized that the proposal was not to form permanent small groups. Jim Isaak suggested that the group might consider the formation of a Heads of Delegation Committee (HODC) to discuss meeting organization, but this suggestion was not taken up. The resolution was redrafted as a result of discussion, and accepted with slight further amendment. 7.2. Action items The action item list appears as attachment 4 to these minutes. Readers should note that it has been reordered relative to a draft circulated at the meeting. A policy decision was taken that pro forma actions should be formulated at this and future WG15 meetings requiring that WG15 memberJbodies, officers and individuals named in resolutions report back on the status of those resolutions at the next meeting. 7.3. Issues list In the following, the notation 9006 refers, for example, to an issue first logged at the June, 1990, meeting. (See also WG15-SD3.) 1. (9006) RIN input to (WG20) Closed. See discussion under J2.2. 2. (9006) Documenting profiles Action has been taken in resolution 117; a response is awaited. The issue remains open. 3. (9006) Canadian comments to 9945-1 Remains open. The issue cannot be closed until 9945-1 is revised or until 9945-2 solves the problem. 4. (9105) Publication of responses to ballot comments New issue. In order that WG15 member's concerns about the visibility of responses to ballots returned to SC22, these responses will be circulated to WG15 as they become available in future. 5. (9105) CD 9945-2.2 document not identified New issue. This was formerly action 9010-7, but was moved to the issues list pending a decision by the US member body that 1003.2 has become stable enough for submission. As well as being required for CD registration, it is also required for liaison with SC2 on character set issues. 8. Closing Process 8.1. Future meeting considerations --JRequest for invitations Jim Isaak presented a summary of meeting location and member body participation since the inception of WG15 work. Randall Howard complained that little could be accomplished between the proposed 1992 meeting dates. Jim Isaak noted that the New Zealand invitation coincided with a break in University term, and so probably could not be moved --Jbut he would check. He also pointed out that three documents could well be in ballot by (northern hemisphere) spring 1992: 9945-2.2; 1003.2A, and 1003.4/.4A, and that ballot results might be available in time for a later rather than an earlier meeting. Ralph Barker pointed out a number of clashing US events in March and April 1992, suggesting a date no earlier than 27th April. John Hill summarized his position on, or more accurately against, meeting in New Zealand. But, since it was the only offer on the table, we should accept it, and not attempt to change the dates. The group agreed that the meeting should be in New Zealand. (See also resolution 151.) After some further discussion, the following meeting locations were agreed upon: November 4-8, 1991 Stockholm, Sweden May 19-23, 1992 Hamilton, New Zealand Late August/early September, 1992 Denmark (in conjunction with SC22?) March/April, 1993 UK (tentative) September, 1993 Germany April, 1994 Japan (tentative) September/October, 1994 Canada (tentative) Delegations from locations shown as tentative above should research the possibilities and report back. The detailed schedule for Stockholm in October is: RIN: 3rd, 4th, 5th AM (possible problem with meeting room on 3rd.) RCT: 4th, 5th AM RSEC: 4th PM, 5th AM WG15: 5th PM, 6th, 7th, 8th, 9th RGCPA will not meet unless notice is given prior to 1st August deadline. 8.2. Document Number assignment The following document numbers were assigned: WG15-N163 Coordination ad hoc minutes WG15-N164 1003.18 WG15-N165 EWOS Guide and cover WG15-N166 Info for participants in coordination ad hoc WG15-N167 1003.0 response to action item on guide as TR WG15-N168 RSEC minutes WG15-N169 RCT minutes WG15-N170 RIN minutes WG15-N171 C language multi-byte support addendum proposal WG15-N172 Ballot comments and summary from "14 ballots" WG15-N173 Programming LIS Methods D4 WG15-N174 NP 1003.3 text WG15-N175 Rotterdam minutes (Attachments: 1) Attendance; 2) Rotterdam agenda; 3) Resolutions; 4) Actions) WG15-N176 Netherlands comments on security (that is, on 1003.6) WG15-N177 First three chapters, 1003.1 LIS WG15-N178 SC22-N963: 1003.4 CD ballot WG15-N179 Agenda and notice for Stockholm 8.3. Selection of Vice Chair & Secretary for next meeting Isak Korn has communicated that he cannot chair the Stockholm meeting because of EC rules. Patric Dempster nominated John Hill in his place. Mr. Hill accepted. There being no other nominations, John Hill will chair the meeting. Martin Kirk nominated Ron Elliott as vice chair. He accepted, and those present concurred. Ron Elliott proposed Martin Kirk as secretary. He accepted. 8.4. Agenda for next meeting Jim Isaak suggested that it would be useful to handle this as suggested changes relative to the current agenda. Under 2.7, there should be a report back from RGCPA. Under 3, the group should review NP progress, 1003.0 and 1003.3. John Hill's report on resolution status should be handled somewhere under 2. 8.5. Thanks to host Herman Weegenaar thanked Jim Isaak for copying services as well as for convening the meeting; Donn Terry proposed thanks to Herman Weegenaar for chairing. Those present wholeheartedly agreed. (See also resolution 159.) 9. Adjournment The meeting closed at 15:58.