Dave I hope the attached file reaches you cleanly. I created it on the Agenda on my flight back, moved it via MS/DOS & Kermit to our Q-Office system to produce a report for my office; brought it back via Kermit to my portable Toshiba at home and did a little editing via Turbo-C compiler(my cheap and cheerful WP on the portable at present!) and then via Kermit to our NCR Tower in order to prepare this mail item for you. After all this interoperability and interworking I hope that it has not lost any text. Working online to UNIX from my Toshiba as I am at present (VT102 emulation) I found some difficulty with Uniplex and screen prints have not been accurate. If it look bad let me know and I will tidy up and resend from the office tomorrow or Friday - its a problem of time at present. Anyway if it looks OK I'll leave it to you to decide whether to e-mail to the POSIX panel Cheers Don Notes from 1st meeting of WG15 Rapporteur Group for Co-ordination of Profile Activity: 20/21 Jan 92 : San Francisco 1. Jim Isaak, convenor of WG15, convened and chaired this meeting, which had fewer delegates than I had expected. Japan, US, and UK provided national rapporteurs. EWOS EG CAE had two reps. Andrew Walker from X/Open carried a WG15/SGFS liaison brief and Roy McKean, also from X/Open, as an observer. There was one other US observer. There is a need for a lead rapporteur to drive the activity and provide continuity between meetings, chairing the meetings and reporting to WG15 as required. Nobody has submitted a name as yet. National bodies are urged to make nomination. 2. The meeting accepted the Issues List from the Profile Co-ordination ad hoc meeting May 91 as an input document. Of the 8 items, 3 could be closed and the others with amendments where necessary were marked as open items on a new issues list. 3. A number of related activities were reviewed, including SGFS, EWOS, OIW, and AOW. The SGFS issues list was reviewed. I encouraged the group not to dismiss user requirements as a topic which belonged elsewhere, suggesting that it was important to try to harmonise the presentation of user requirements, citing the work of DISC as useful in this respect. The DISC FUR was tabled as a review document, and I was requested to forward a copy to IEEE TCOS Profile Steering Committee. Willem Wakker provided the EWOS CAE review. It was noted that OIW now is scoped to cover OSE. The body should provide a liaison to this Group. The Japanese representative said that his Government, through MITI, was trying to decide whether AOW, SIGMA, or the National Standards body with an extended brief should handle OSE. A decision is expected by April. As SIGMA was set up to handle OSE issues from a software development perspective, it is a favourite candidate. MITI has also initiated a study of Japanese Government procurement. 4. National rapporteurs were requested to invite profiling bodies in their countries to participate in this Group. For the UK I was requested to write to CCTA and X/OPEN initially. 5. Terms of Reference The following Terms of Reference were proposed (subject to approval by WG15 at its May 92 meeting): (i) To co-ordinate the participation of WG15 in all profiling activities, including, for example, framework and taxonomy development, related to WG15 standards. (ii) To identify how WG15 standards should be used in profiles, thus promoting co-ordination and harmonization of those profiles, and to provide guidelines for the further development of WG15 standards in order to ease their use in profiles. (iii) To monitor activities relating to SGFS activities and profiles referencing WG15 standards, and to provide feedback directly to SGFS on both procedural matters (that is, use of the profiling techniques) as well as technical content of profiles 6. Programme of Work The following programme of work is proposed: i) Identify, establish and maintain liaisons with groups developing, or planning to develop, profiles. ii) Invite developers of profiles referencing WG15 standards to participate in the RGCPA for the purposes of co-ordination. iii) Review of profiles: - Review and make recommendations on the IEEE P1003.0 guide. - Obtain early visibility of profiles that reference any POSIX standards(not limited to WG15 work items). - Review and respond to SGFS on behalf of WG15 regarding ISP and TR10000 issues. - Advise WG15 about new work required to enable current or future profile production. 7. TR10000 We looked briefly at TR10000-1.3, a revised draft framework and taxonomy for OSE. Topics that need further consideration are, conformance, use of extensions, acceptance of gaps. It is noted that the OSI world has made free with the alphabet in selecting single letter identifiers for classes of profiles. The OSE community will have to fit around the existing F,A,B,T,U,R types. P is likely to be selected for OSE profiles and, as I observed it would be conceivable for a P profile to incorporate an F,A,B,T,U,R profile. We should look at section 7.2 of TR10000.1-3 and identify similar classifications to 7.2.1 for 7.2.2. I tabled a paper from the IEEE TCOS PSC which offers a technical sub-structure based on P1003.0, the order of services descriptions in Sections 4 and 5 of the Guide to the POSIX OSE. 8. FIPS 151-2 It is proposed to compare FIPS151-2 with the P1003.18 profile, the objective to identify common elements which I suspect will be entirely within the IS9945-1 domain. The aim is to maximise harmonisation. Options in P1003.18 will be contained in an Annex. Donn Terry would welcome any comment on this profile. 9. P1003.0 We looked at the outline timetable for this Guide en route for ISO TR. It was observed that it will be difficult to deliver precision on standards references in view of the lead time in the standardization process, ie when the Technical Report is published many references under 'emerging standards' might well have reached full international status. We need to ascertain whether there could be a faster process, eg fast track to deliver such a TR. An amendment process initiated fairly soon may be necessary to accommodate changes to standards references. The IEEE P1003.0 committee should discuss this issue. 10. Test Methods This topic has had an airing at the IEEE TCOS PSC but no firm opinions are formed. Is it essential for conformance tests for profiles to include the base standards test? Are there new tests required to test the elements that bond the base components of the profile? It was agreed to ask WG15 RGCT (conformance testing rapporteurs) to look at the issue and prepare a liaison statement to deal with conformity clauses in base standards in respect of profiling. A 3 level conformance structure was noted: strictly conforming, (where an application requires only the facilities described in IS 9945 and the applicable language standard), - pure but perhaps not very useful conforming, (where an application uses only the facilities described in IS9945 and approved conforming language bindings for any ISO standard), more useful conforming with extensions (where an application uses non-standard facilities that are consistent with IS9945.) The latter may have two manifestations, overlapping (eg, TCP/IP), and non-overlapping (eg, Windows) with existing standards. It was observed that APIs are quite a different beast from the real interfaces of OSI. Testing of APIs brings more problems. 11. Next meeting will be in October, back to back with WG15, precise date yet to be decided.