-------------------------------------------------------------------- ISO/IEC JTC1 SC22 WG16 N188 Minutes of 15th Meeting of ISO/IEC JTC1 SC22 WG16 LISP, November 8, 1996, Bath, United Kingdom. -------------------------------------------------------------------- Source: WG16 convenor Action: Information -------------------------------------------------------------------- Meeting record by Julian Padget 1. Opening of the meeting ========================= PP welcomed the participants to Bath and thanked JAP for hosting the meeting He also informed us that the new secretary for WG16 at AFNOR is Arnaud Diquelou, email: diquelou@orion.afnor.fr, tel: +33 1 42 91 57 18. 2. Roll call of delegates ========================= PP: Pierre Parquier (convenor and FR HoD) WW: Whitman Wright (CD HoD) TI: Takayasu Ito (JP HoD) TY: Taiichi Yuasa (JP) RJB: Russell Bradford (GB) observer JAP: Julian Padget (GB HoD) 3. Adoption of the agenda ========================= Recognition of distributed documents N180 Canadian proposal for graphics in ISLISP N181 WG16 report to 9th SC22 plenary N182 WG16 15th meeting agenda N183 Table of replies on DIS 13816 ballot N184 ITTF comments on DIS N185 Canadian proposal for graphics primitives in ISLISP 2000 Item 9: addition of discussion of N181 Item 10: addition of document reference. Discussion of ITTF comments Item 11: addition of document references, including discussion of N185 Agenda adopted. 4. Appointment to drafting committee ==================================== All present will assist in drafting, then JAP will pass them on to KMP. 5. Liaison reports ================== No liaison 5.1 Reports from national member bodies ======================================= CD: expressed pleasure at the progress in the work of WG16. Have made a proposal for graphics extensions to be discussed later. Anxious to see implementations of ISLISP. JP: two meetings. First shortly after Paris meeting at which delegates reports on Paris meeting. Second discussed how to respond to DIS ballot. GB: no meetings. No e-mail discussion and hence no comments on DIS. FR: two meetings: one reporting and a second to prepare comments and vote. 6 Review of priorities and target dates ======================================= We have to issue a Disposition of Comments based on N183. (addressed later in item 10) When the Disposition of Comments is received by ITTF they then collaborate with the project editor to correct minor editing bugs. Finally the IS should be published. Once published, users of the IS will issue defect reports to ITTF. WG16 should define who these defect reports should be forwarded to. RESOLUTION 1: The committee hoped that PP and KMP would handle defect reports, but should either of these experience a change of circumstances, JAP and TY volunteered to be available to take over from one or the other as appropriate. Resolution approved. 7 Work item on which no progress is being made ============================================== None 8 Review of Project Editor and liaison arrangements =================================================== The project editor sent the DIS to ITTF for the DIS ballot. Ballot closed on September 23, with no coments on the contents. ITTF sent a few comments to the project editor on minor editing bugs. RESOLUTION 2: WW noted that given the size of the document the list of ITTF comments is very short and proposed a vote of thanks in recognition of the excellent work that KMP has done in preparing the DIS/IS. Approved unanimously. The ITTF comments were noted and the progress towards their implementation was observed in the documents supplied by the project editor. Some items are outstanding. Project editor will send revised document to ITTF for publication as an IS as soon as these changes are completed. 9 Review of recent JTC1 decisions affecting the group ===================================================== Refer to N181. No request from WG16 to SC22, and no particular comments from SC22 to WG16. Work management: ISO is re-engineering itself and examines every project and ask each NB whether the project is of interest: in effect a second new work item ballot. At least 5 NBs must commit to support of a project. WG16 has received sufficient support. DIS ballot is complete, resulting in 21 approvals, 5 abstentions, and no vote against. 10 Preparation of Disposition of Comments to DIS 13816 vote =========================================================== TY: already addressed by ITTF in N179. NL: already addressed by ITTF in N179. FR: we accept the comments and request the project editor to proceed with their implementation. To be document N186 - approved unanimously. 11 Items of future work ======================= Topics raised: graphics primitives (N185), modules, internationalization, bindings. 11.1 GRAPHICS ------------- CD presented N185 which was prepared for FORTRAN 2000. Part 1 contains general remarks, Part 2 contains specific proposals (eg. names of routines and descriptions of behaviour), Part 3 contains commentary on and some revisions of the earlier material. FR: drew attention to the fact that over 10 years ago, Le-Lisp followed the raster graphics route; feel that GKS is out-dated for general purpose. Also, general purpose graphics requires higher level oparations (eg on the Web, AWT replaced CGI, and is still too simple). Instead of deciding for a single philosophy (vectorial or raster) and a single level of complexity, FR would prefer a binding based solution. GB: observed that although line-drawing capabilities may be suitable for FORTRAN applications, the preponderance of ISLISP applications may be more raster-oriented. CD: does not feel that the computer industry has delivered suitable products for professions where the market for individual applications is small as compared to business. JP: appreciate position and agrees we should have something in the future, but it would be hard to define something suitable now, both in terms of community needs and in terms of the maturity of graphics standards. GB: don't believe there is a single good solution. Find it more attractive to be able to build an API to some chosen package. CD: for the sake of portability of ISLISP applications with graphics, many business applications are willing to sacrifice some level of efficiency and functionality. They are willing to accept a single reasonable package. FR: suggested defining a separate standard rather than an appendix. CD: welcome more comments through islisp mailing list. 11.2 MODULES ------------ JP: reported that have an implementation of the scheme described in N159 for an interpreter, currently working on compiler. Based on our experience, we might come up with a modified proposal or rationale. JAP: different groups will have radically different requirements for modules, contrast Ilog Talk, EuLisp, Japanese module system and the various Scheme proposals. RJB: modules are a good idea, but major complications arise from interactions with macros Both PP and JAP expressed varying degrees of dissatisfaction with their existing implementations. RESOLUTION 3: we are not ready to address the specification of a module system at this stage. The meeting agreed that we should let the members of the working group gain more experience with their different schemes before proceeding with this issue. Passed unanimously. 11.3 INTERNATIONALIZATION ------------------------- TY: talked to WG20 representative who had advised waiting for outputs from that WG about character set support in programming languages, instead of moving ahead independently. Thus, the working group concluded that it would be better to wait for guidance from WG20 with regard to character sets. No interest was expressed for other aspects of internationalization. 11.4 BINDINGS ------------- The discussion identified three issues (among many) that were felt to be particularly important: 1. binding from ISLISP to C 2. binding from ISLISP to C++ 3. binding from C/C++ to ISLISP from which follow several technical issues o implications for garbage collection o conflict between C++ and ISLISP exceptions o generic function dispatch o cross-language subclassing o multi-threading TI: Identified four categories: fusion, integration, binding and interfacing. C is binding, but C++ demands integration. Interfacing just means linking things that are already compatible. Concluded that a C binding should be attempted before doing anything more ambitious. JAP: to send the EuLisp to C binding to the WG. RESOLUTION 4: to start work on defining an ISLISP/C binding to be included in a subsequent revision of the standard. Approved unanimously. 11.5 PROMOTION OF ISLISP ------------------------ Dave DeRoure (DDER) was supposed to prepare a WWW page for WG16. JAP will follow up. Should include references to existing implementations and IS publication date. PP to send IS publication date and French implementation reference to JAP/DDER. Action item for each member of the WG is to advertize the web page when it is available. GB must announce availability of WWW page (deadline Jan 1 1997). CD: ISLISP should be made available for academia JP: it is necessary to advertize ISLISP to LISP communities and AI communities, and after IS is published it is hoped that ISLISP implementations will become soon available in japanese academia with collaboration of JP manufactures; an international collaboration would be useful to have good ISLISP implementations. 12 Subsequent meeting requirements ================================== No further meeting is required yet. We shall discuss future meetings on islisp mailing list when the need arises. 13 Any other business ===================== None 14 Approval of resolutions ========================== All approved unanimously. --------------------------------------------------------------------