Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!usc!orion.oac.uci.edu!ucivax!gateway From: JPALME@qz.qz.se (Jacob Palme QZ) Newsgroups: comp.protocols.iso.x400 Subject: Notes from the ISO/IEC/CCITT X.400 development group Message-ID: <525662*JPALME@QZ.qz.se> Date: 25 Nov 90 17:34:08 GMT Lines: 358 Approved: usenet@ICS.UCI.EDU In-Reply-To: <525660*JPALME@QZ.qz.se> Autoforwarded: true Here is the complete text of the minutes which I wrote during the discussions in the computer conferencing/group communication subgroup at the ISO/IEC/CCITT meeting. SOURCE: Joint ISO-IEC-CCITT Group Communication subgroup at the co-located meeting of CCITT Study Group VII/Q.18 and ISO/IEC JTC 1/SC 18/WG 4 SPG on Messaging, Geneva, November 1990 (Jacob Palme, notetaker) TITLE: Notes from the Group Communication Subgroup STATUS: These notes mainly present the discussions within the subgroup, not the final decisions made by the group. The final decisions made by the group are recorded in a separate status report from the subgroup. Wednesday, November 14, 1990 ============================ List of participants Steve Benford, University of Nottingham, Nottingham, United Kingdom NG72RD (Chairman of the group) Michael Durkin, Bell Atlantic, One Parkway, 9B, Philadelphia, Pa 19102, U S A. Ank Hoang-van, France Telecom Cnet, 38/40 General Leclerc 92131 Issy Les Moulineaux, France Michael Kuna, Herrenhausstrasse 10, D-1197 Berlin, Germany Hiromiki Moriyama, NTT-PC Com, 3-23-11 Nisahi-shinbashi, Minato-ku, Tokyo 105, Japan Jacob Palme, Skeppargatan 73, S-115 30 Stockholm, Sweden Jean Walravens, Alcatel Bell Telephone, Francis Wellesplein 1, B- 2018 Antwerpen, Belgium Ulrich Zysset, Swiss Ptt/KS 3, Viktoriastrasse 21, 3030 Berne, Switzerland List of contributions TD 4019: Meeting report from the GC subgroup in the previous VII/Q.18 meeting in Munich. TD 4020: Draft recommendation F.gc, version 3 (output of the Munich meeting) TD 4021: Draft recommendation X.gc, version 3 (output of the Munich meeting) TD 4022: GC issues and requirements, version 3 (output of the Munich meeting) TD 4097 Asynchronous Group Communication, liaison letter from SG I. This letter has two appendices: 1. Draft recommendation F.agc, version 4 2. Functional specification for a typical computerized conferencing application D-210: Draft recommendation X.gc, version 4 (from Sweden) D-211: Draft recommendation F.gc, version 4 (from Sweden) D-269: Comments on the standardization of Group Communication (from NTT, Japan) D-300: Group Communication Information Model and Abstract Service (from UK) D-308: Direction for Group Communication (from UK) MH-300: Different Group Communication Implementations (from Jacob Palme, Sweden) MH-301: Contribution on Draft Recommendation F.gc/X.gc (from Michael Kuna, Germany) GC-1: Why Attribute Carriers (from Jacob Palme, Sweden) Planning of work during this meeting 1. Identify output and timetable for this study period (Input paper D-308) 2. Relation to other work (Input paper 4097) 3a. Technical issues, information model (Input papers D-210, D-211, D-300, GC-1) 3b. Technical issues, architecture (Input papers D-269, MH-300, MH-301) 4. Issues reflecting on the MTS (restriction on distribution) 5. Identify what we need for the next meeting Resolution of different requirements on the GC work We noted that there are two different and possibly conflicting views on the output wanted from the group communication work: a. That we produce a general-purpose group communication model and framework, which can form as the basis for different group communication applications. This general model and framework should not be restricted to modelling group communication as part of MHS. b. That we produce, if possible within the present study period, a recommendation/ standard for the interconnection of computer conferencing system, which has a close relation to MHS. We made a tentative decision to resolve this problem by planning to produce three documents before the end of the present study period: Part A: Service specifications and user functionality (corresponding to the F.gc document) Part B: A general framework document. This part can contain: > A general information model > A general architecture model > A general management model > A general description of user operations Part C: Computer conferencing. This part can contain: > How to apply the models in document A on computer conferencing > User operations in computer conferencing > The specific model, architecture and protocol for computer conferencing Planning of future work during the present study period Our goals (as seen at this stage of the subgroup meeting) would then be to produce, during the present study period, Part B to CD and CCITT draft recommendation status, Part A and C to CD and CCITT recommendation status. The reason why part B would not go to DIS status is that we expect that the 1992 CCITT recommendation will be restricted in scope and functionality, and that ISO will want to extend the scope and functionality together with CCITT in the next study period. (See the status report for the final goals agreed at the end of the subgroup meeting.) Discussion of other work We noted that work on synchronous group communication is presently being done in SG I. We accepted that our results in part C (computer conferencing) will be mainly aimed at asynchronous group communication. We decided to write a liaison letter to SG I presenting the above plans and asking for their approval. We also decided to ask SG I for liaison from the present work on synchronous group communication. First pass at discussion of information model We discussed the information model, mainly as presented in the input papers D-210, D-211, D-300 and GC-1, noting the similarities and differences in view on the information model. Thursday, November 15 ===================== Future plans Carl-Uno Manros visited us and said that there is no requirement that we must be ready with a standard within the present CCITT study period, but that he does require us to decide, before the end of this meeting, whether we plan to have something ready for approval as a CCITT recommendation within the present study period. We decided to postpone a decision on this until Knut Smaaland can join our group next week. Restricted domain We decided to write a letter to the Q.18 plenary, asking the plenary to transfer responsibility for the function "restricted domain" from us to the general enhancements subgroup. Jacob Palme will draft a letter. What kind of standards should we produce We had a long discussion on what kind of standards we are to produce. One alternative would be to produce the standards document in three pieces: A. A general definition of a group communication information model, general architectural model and abstract service specification. B. A definition of computer conferencing, its abstract service specification and how it can be realized using the general model produced under A. above. C. A realization of the general model produced under A., as it can be used for computer conferencing, using partly existing services like MTS, DS and DFR. Another alternative would be to only produce the document B and C as listed above, with direct specification of how B can be realized using C. A third alternative would be to specify A, B and C, but to define the mappings not only from B to A and from A to C, but also between A and C. We made the tentative decision to try to develop this, but to be prepared to skip the A. model if it is found to difficult or unsuitable. Plans for work during the present meeting We decided to plan our work during the present meeting as follows: Thursday afternoon: Computer conferencing model. Friday morning: General GC model and mapping of the computer conferencing model on it. Monday: Summary of work during the previous week and work on the realization using (wholly or partly) MTS, DS and DFR. Tuesday: Concluding work, editing. Computer conferencing model Operations: Role: Member Subscription-request/Unsubscription-request (on an activity, a conversation or a conversational branch). Find-activities Find-subscribed-activites Find-contributions (includes finding contributions which are unread, which are replies, directly or indirectly, to another contribution etc.) Fetch-contribution (in order to read it, print it, file it, filter it etc.) Re-submit a contribution (not necessarily self-written, to another activity) Add/remove keyword from a contribution Find-members of an activity Lookup information on another participant Submit contribution Reply to all recipients of a previous contribution Reply to only the author of a previous contribution Write a contribution obsoleting a previous contribution Delete a contribution Moderator operations Pre-moderation: Find-submission Fetch-submission Accept/reject submission (whether or how to notify the submitter is FFS) Re-route submission (to another activty) Post-moderation: Remove or reroute contribution (may include automatic removal or rerouting of the conversational branch started by the contribution) (Note that removal does not mean physical removal of the contribution text itself, since the contribution may still be available e.g. in other activities or to its author) Owner/General manager operations Set access controls. Decide which entities or roles are allowed to copy, reroute or remove contributions from an activity. Create activity and sub-activity. Delete activity (This may not imply physical removal of the contributions in the activity, since they may be available as contributions in other activities and to their writers). Modify-activity (Some, but not all, attributes which can be modified are expiration time, access controls, description, name, aliases, keywords, rules, encryption policy, suspension status). Add/remove members from activities. Add/remove owners from activities. Register and deregister users. Notification handling At least he following notifications may be needed: Notification of acceptance and rejection of submissions to pre- moderated activities. Notification that you have become a member of an activity. Delivery notifications to the conference and to its members. Receipt notification. Issues for further study in the above operations list: Notifications and their use Naming scheme Handling of anonymous or pseudonymous contributions How to announce new activities to potential members Filtering Setting keywords on activities Friday, November 16 =================== On Friday we discussed the general group communication model. Steve Benford presented his model. The general opinion was that we should develop such a model, but that such development was a long term task, and that we would in parallel develop a computer conferencing abstract service definition and implementation based on MHS. We also decided to produce liaison letters to SG I (regarding future work on the F document) and to Q.20/ISO SC21/WG 4 on use of the directory for group communication. (Later in the meeting we decided not to produce any liaison to the Directory Group yet, since our work on how to use the DS is not complete enough to present to them.) Monday, November 19 =================== On Monday, we discussed the Computer Conferencing Information Model, Abstract Service definition and Architecture. The results of this discussion can be found in our working paper, X.gc version 5, part 3. Important issues during the discussions on Monday were to what extent Computer Conferencing is to rely on the existence of a globally interconnected Directory System and which underlying services are best used for communication between GCSA-s. We then decided to produce the following output documents from this meeting: Document Who will prepare a draft -------- ------------------------ X.gc version 5 Jacob Palme with input from Steve Benford GC subgroup notes Jacob Palme Liaison letter to SG I Jacob Palme Restricted Domain proposal Jacob Palme Timetable and status report Steve Benford We then made the following table of documents which we would like to see for the next Q.18 meeting: Document Who is willing to work on this -------- ------------------------------ Abstract service for computer conferencing Jacob Palme, Steve Benford and including ASN.1 encoding Michael Kuna Distributed architecture Jacob Palme and Michael Kuna for GC General GC Information Model Steve Benford, Michael Durkin and Jacob Palme Computer Conferencing Steve Benford, Michael Durkin Information Model and Jacob Palme We decided to radically change the text in X.gc version 4 to show how computer conferencing operations and architecture can be derived in a top-down fashion from the computer conferencing abstract service definition. Since we did not have time to finalize this work during the present meeting, we will not copy the old text of X.gc into the output working document, even though parts of the ideas in this document will have to be handled in the final X.gc document. Tuesday, November 20 ==================== Tuesday morning was spent on the general information model. We especially discussed whether links are objects themselves or only relations between objects. On Tuesday afternoon, we went through a paper with functional requirements on group communication coming from the the EIES development group in New Jersey. We checked their requirements against our requirements, and added proposals for new requirements where needed.