Unapproved Meeting Minutes BSI IST/5/-/15 (Posix) Panel Date: 28th May 1993 Location: Hampden House, London Those Present: Dave Cannon (Chairman) University of Exeter Andrew Walker X/Open Martin Kirk X/Open Don Folland CCTA Peter Owens CCTA Ross Allen BSI QA David Brownbridge Praxis Derek Jones Knowledge Software David Rogers Data Logic Neil Keskar Computer Sciences Corp Papers before the meeting: P210 Agenda: IST/5/-/15 meeting, 28-May-1993 P211 Unapproved Minutes, IST/5/-/15, 26-Feb-1993 P212 WG15 Actions required of UK P213 WG15 N375: Secretary's Action Item report P214 WG15 Meeting report, Heidelberg, 10-14 May P215 Verifying Claims of Open Systems Conformance P216 IEEE PASC Ballot Group Invitations P217 IEEE PASC Posix group status report P218 IST/-/2: Request for input on revision of TR 10000-1 & TR 10000-3 P219 WG15 RGCPA: Notes, 10/11 May M01 Addresing the Posix LIS issue: Martin Kirk The chair thanked Dave Cannon for producing the mailing for this meeting and reported that Dan Chacon had agreed to take this on for future meetings. 0. Apologies Apologies were received from: Dan Chacon Colin O'Driscoll. 1. Adoption of Agenda P210 The agenda for the meeting was agreed as presented. 2. Approval of Minutes P211 The minutes of the previous meeting were approved, with a correction to the spelling (David Brownbridge), and to status of the C Addendum (Working Draft). 3. Matters Arising Colin O'Driscoll had forwarded the NP on 'Conformance Testing Methodology' to Jim Isaak as requested, as the UK submission. Action: Dave Cannon agreed to find out whether David Ferbrache would be willing to be the IST/-/15 rapporteur to RGSEC. (Done: 1-Jun-1993) Action: Chair to contact Greger Leijonhufvud, discover where he is employed now and whether he attended the SC22 Ad-Hoc on Charactersets. (Done: 29-May-1993) Action: Chair to contact GL re the WG14 ('c' Group) meeting in London from 19th to 21st of July. If available could Greger attend to discuss stateful encoding problems? (Action from February Posix Panel). (Done: 1-June-1993) 4. Progress of ISO/IEC JTC1/SC22/WG15 Dave Cannon reported on the Heidelberg meeting, which had been attended by himself, Martin Kirk and Don Folland (for RGCPA), together with Dan Chacon and Colin O'Driscoll (for RGCT). At the April POSIX meeting the SEC had decided to withdraw the requirement for Test Methods. WG15 approved the drafting of an NP request to SC22 to permit the standardization of Test Method drafts coming from POSIX. LIS is still technically in the melting pot because the US TAG requested at the SEC that they be given time to consider alternatives to the PASC producing LIS drafts. One option would be to find another body to produce LIS drafts, such as ANSI, but the coordination problems would be considerable. The problem of manpower will remain the same. Another option would be for the work to be done in ISO. The general view is that at its next meeting IEEE will confirm that LIS should not be required. There had been discussion of the French comments on the overlap and potential conflict between Ada Real-time work in WG9 and the Ada Binding to 9945-1:1990. M01 Martin Kirk introduced a proposal for a UK position on the timing of LIS standards. It suggested that an LIS specification should not be required until a second language binding was needed. The benefit would be that it would allow standards development to continue whilst still keeping in place the requirement for LIS. David Brownbridge expressed concern that this proposal went beyond the discussions at the previous meeting, and that in the end the result might be either that LIS specifications would never be produced, or that there would be a rush to create standards based on specific language bindings. Derek Jones suggested that LIS specification development could wait until the first revision of the standard. There was general discussion on the nature of conformance to a LIS standard, and how LIS definitions should be tested. It was argued that the real requirement is conformance to LDS specifications, not LIS. Dave Cannon pointed out that the previous position of the group had been that in the short-term Language Dependent Specification (LDS) definitions should be accepted, and that this should be reconsidered by WG15 when the work of WG11 had reached fruition. There was a general conclusion that the current LIS approach was not adequate to the need, and that what was needed, if anything, was an abstract definition of the underlying data and functionality. Don Folland said that his concern had been COBOL, but that in practice COBOL programmers were making adequate progress using a C Binding. It was agreed that the proposal from Martin should be amended to say that there should be no initial requirement for an LIS definition, and we would not expect to visit the LIS problem until either WG11 completes its work, or a second language binding appears, whichever is sooner. Action: Martin Kirk to redraft LIS proposal (M01) as per minutes, and pass to Chair for submission to IST/5. 4.1. Consideration of ISO Ballots 9945-2 closes 6/6 Dave Cannon had passed a submission to the IST/5 secretary, voting YES on the 9945-2 draft following 'phone consultations with interested Panel members. There was discussion on the need for line numbers, and it was decided to make a late submission asking for them, starting at 1 at the top of each page. Action: Chair to contact WG15 Technical Editor commenting that 9945-2 should include line numbers. (See later action) 4.2. Status Report of SC22/WG15 P214 Rigorous Conformance WG15 N374 ('Rigorous Conformance' requires applications to conform to language requirements as well as service definitions). Derek Jones asked what had happened to this paper. Dave Cannon reported that he had given copies to John Hill and to the POSIX .1 group. It is now an official WG15 document, and will be considered by the US TAG. Derek pointed out that two UK Government procurements were asking for proof of conformance to POSIX by an application. Action: Dave Cannon to follow up the Rigorous Conformance issue in POSIX 1003.1 and WG15. Real-Time For ISO the .4 and .4a work was being merged, and Dave Cannon reported that the Disposition of Comments meeting on the 'Real-Time' amendment to 9945-1, held in January, reached a satisfactory conclusion on all the comments and WG15 anticipates that the SC22 secretariat will pass the resulting document for DIS circulation. Actions from Heidelberg P212, P213 9305-01 UK National Body to further scope the work required as defined in N347 Action: Colin O'Driscoll: scope the work as defined in WG15 N347. 9305-08 Canada, UK: Nominate candidates for the position of Liaison to ECMA TC33 (open action item 9205-59, 9210-17) Action: Don Folland: check possible candidates for WG15 liaison to ECMA TC33 9305-28 Member Bodies: Review N375 to determine status and report back to next WG15 meeting. (WG15 N375 available as document P213) The meeting identified a number of items from P213 which were relevant to the UK. Par.007 Member bodies to contribute national Profiles that they have developed to WG15 RIN and to WG20. Action: Greger Leijonhufvud: pass any UK profiles derived from meetings with Welsh experts to WG15 RIN and WG20. Rot.003 WG15 member bodies to review and comment on the taxonomy in the EWOS Technical Guide on Profiles for the OSE. (EWOS/TA/81/68, WG15 N165) Action: Peter Owens: pass any comments on the taxonomy in the EWOS Technical Guide on Profiles for the OSE. (EWOS/TA/81/68, WG15 N165) to WG15. Rot.004 WG15 member bodies to review and comment on IEEE P1003.18 (January 1991), WG15 N164. No longer relevant. Rot.005 WG15 member bodies to review and comment on the final report of TSG-1. No longer relevant. 9305-42 Member Bodies: Member bodies having comments on 9945-2 are requested to supply them via E-mail to Project Editor as well as the official ballot return, in order to give as much resolution time as possible. Action: Dave Cannon to email Hal Jespersen asking for line numbers as the UK comment on 9945-2. 9305-45 Member Bodies: Review the following general specification methods with regard to the requirements listed in Resolution 93-233, and provide a written response prior to the October 1993 WG15 meeting, evaluating these methods for future use in projects JTC1.22.21, JTC1.22.39, JTC1.22.42. Member bodies are invited to suggest and evaluate additional specification methods. 1. LIS methods described in N173 (TCOS-LIS) 2. LIS methods under development by WG11 (LID, LIPC) 3. LIS methods using formal definition techniques (VDMSL, Z, others) 4. Single binding document with multiple language syntaxes 5. Seperate thick language bindings Action: All panel members to consider the future requirements for LIS support in WG15 standards. IST/5/-/15's interim position is to welcome LIS drafts written to comply with WG15 N173 (TCOS-LIS) (Option 1), but to accept drafts written as separate thick language bindings (Option 5) where LIS versions are not available. IST/5/-/15 currently recommends revisiting the problem when WG11 completes its task. (Option 2). This issue is for consideration at the next meeting of IST/5/-/15. 9305-50 Member Bodies: Invite the appropriate experts involved in conformance testing to participate in RGCT representing their respective countries. Action: Dave Cannon to phone James Andrews at X/Open (0734 508311), a NAMAS contact, to check for any interest in RGCT participation. (Done: 2-Jun-1993) [JA said he'd talk to David Raynor (NAMAS, based at NPL) to see if he had any interest in participating. DR to contact me.] 9305-61 Member Bodies: Report any issues, concerns on advancement of POSIX Profiles internationally to WG15. (Resolution 93-232) Action: Martin Kirk to ask within X/Open and Don Folland to ask within CCTA whether either organisation had any concerns re the advancement of Posix Profiles, and to report back. 9305-64 Member Bodies: Provide names of delegates for both WG15 meeting and WG15 Rapporteur meetings to the convener of the meeting before the start of the meeting. (Resolution 93-242) Action: Dave Cannon to advise WG15 convener of delegates prior to the start of the WG15 meetings. (Continuous...) 9305-65 Member Bodies: Host for spring 1996 WG15 meeting is required. Volunteers are requested to supply an official invitation to the convener. Action: Panel members to consider possibility of hosting WG15 meeting in the spring of 1996 and advise. (Continuous...) 5. Progress of IEEE TCOS P1003 P216 Action on ballots was up to individual panel members as appropriate. P217 David Rogers reported that the PAR for P1003.22 is now approved, and the group is producing a Guide to POSIX Open Systems Security Framework. P1003.6 has a new chair, Lyn Anbuel, and the .6 draft is in ballot resolution; they are looking at potential new areas of work in authentication, graphic interfaces and system administration. P1003.21 has also been approved: its scope is Real-Time Communications. Martin Kirk reported that P1003.7.1 is going into ballot in June. A PAR has been submitted for User and Group Administration. Don Folland reported that P1003.0 was supposed to submit draft 16 to Heidelberg, but it is now expected in July/August. This delay is due to the P1003.0 Technical Editor's funding being frozen during the transition to the Clinton administration in the USA. 6. Rapporteur Groups Security - no rapporteur. Verification - no report from Colin O'Driscoll. Internationalisation - Greger did not attend. P219 Profile Coordination - Don Folland reported. There was little to coordinate, and it would be difficult to do any work until the October meeting. RGCPA will meet again at the next WG15 meeting, in the US, October 25-29. Don reported that he may not be able to continue as UK rapporteur, and felt that there would be benefit in a User representative doing the work. 7. Liaison Reports Dave Cannon reported on the last IST/5 meeting, which had accepted the view of the panel on LIS as an interim UK position. P215 Derek Jones reported on the meeting of the C panel. WG14 is creating a defect reporting log. C will be the first language standard to support multi-byte characters. This work is different from the X/Open specification. X3J11 has a subgroup which is working on numerical extensions. The Panel discussed Derek's P215 paper on 'Verification of Claims to Open Systems Conformance'. Action: Panel members are invited to send comments on P215 to Derek Jones. P218 Document P218, Request for comments on the UK position to SGFS, was discussed. (Derek Jones reported that the X/Open POSIX XPG4 profiles used the same terms as POSIX but with different meanings). Derek also commented on P218 that the work had taken inadequate note of POSIX work, especially .0 and .3. Action: Dave Cannon to phone Teddy Ngosi, saying that the panel is happy with the concept of the paper, but point them to P1003.0, the .1 LIS, and .4. The panel may be interested in the full text of TR10000-3 when it is revised. (Done: 2-Jun-1993) Action: Dave Cannon to send current version of TR10000-3 (if available) to Derek Jones. (Done: 1-Jun-1993) SGFS - Andrew Walker had not attended a recent IST/-/2, and was not planning to go to the next SGFS, so no report. 8. Any other business Next Meeting 10th September 1993, 10:30 am, Hampden House.