Unicode Technical Committee Meeting #64 --------------------------------------- Friday, March 10, 1995 Cupertino, CA (Hosted by Microsoft Corproation) Chair: Joan Aliprand Vice-Chair: Mark Davis Secretary: Ken Whistler Attendance: ----------- Full Members: Andy Witkowski, AT&T GIS Peter Edberg, Apple Tim Greenwood, Digital Equipment Co. Ecological Linguistics (Glenn Adams proxy for Lloyd Anderson) Tae-Jin Kang, Hangul and Computer Co. Ltd. Don Carroll, Hewlett-Packard Mike Ksar, Hewlett-Packard Alexis Cheng, IBM (proxy for John Gioia) Lisa Moore, IBM Ed Batutis, Lotus Lori Hoerth, Microsoft Sun Gi Hong, Microsoft Mike Kernaghan, Microsoft Murray Sargent, Microsoft Rick McGowan, NeXT Lloyd Honomichl, Novell Joan Aliprand, RLG Ken Whistler, Sybase Mark Davis, Taligent Arnold Winkler, Unisys Liaison: James Do, TCVN/JTC1 Officers: Joe Becker The UTC meeting was called to order by the Chair at 1:10 pm. New members ----------- Hangul and Computer Co. (represented by T.J. Kang) and Sybase (represented by Ken Whistler) had been recognized the previous day, during the Editorial working day. Approval of UTC #63 Minutes: ---------------------------- Arnold Winkler requested that his e-mail address be corrected. Motion moved by Honomichl, seconded McGowan: That the Minutes of UTC #63 be approved as amended. Vote: Yes 13, No 0, Abstentions 2, Absent 0. Calendar in Review: ------------------- UTC #65, on June 2, 1995, will be hosted by Novell, either in Sunnyvale or San Jose. UTC #66, on September 29, 1995, will be hosted by Lotus in Cambridge, MA. UTC #67 will be held in the Bay Area. Due to time constraints, the request by Ed Hart for a change in date (to allow for co-location with the X3V1 and X3L2 meetings) was tabled (deferred) until UTC #65 (June). The remainder of the agenda was rearranged as follows: 1. Book schedule; 2. Closed Caucus 3. Conformance (semantics and properties; UTFs) 4. All other editing issues: Equivalency; "Combining" vs. "non-spacing" terminology; Properties; Vietnamese combining tones; Deprecate 343 koronis; Remove reserved range for R-L scripts; Half forms; Chenge rules for ZWSP so that it affects rendering; Widths (half and full). 5. Instructions to the WG2 Liaison 6. New scripts (Tibetan, Ethiopian) 7. General editing Book Schedule: Discussion concerned the problem of publishing the new edition of the Unicode Standard before assignment of code points by WG2. Motion moved by Edberg, seconded by Cheng: That the next published version of the Unicode Standard will not include new or modified codepoint assignments unless they have been approved by ISO/IEC JTC1. Discussion centered on the general issue of whether the Unicode Consortium can publish new draft codepoints prior to approval by ISO. Those opposed to use of draft codepoints advocated use of the Private Use Area for characters without codepoint assignments in the main code space. A friendly amendment that it is ok to include a repertoire but not any code point assignments was accepted by Edberg, but was subsequently withdrawn. Vote (on the unamended motion): 5 yes, 10 no, 0 abstentions, 0 absent. Motion failed. CD-ROM Contents: ---------------- The new edition of The Unicode Standard will be accompanied by a CD-ROM containing mapping tables, etc. The information is being collected, and needs to be reviewed. Action Item #1-Mark Davis: Send out e-mail requesting additional material for the CD ROM. Action Item #2-Rick McGowan: Make a separate directory in the private part of the FTP Site for the CD ROM material. Notify Corporate Members of the location of the information. Lori Hoerth (Microsoft) volunteered to be editor for the mapping tables. The meeting entered Closed Caucus. (Separate Minutes are sent to Corporate Members only.) The open UTC resumed at 3:47 pm. Sargent had to leave. Publication Schedule: --------------------- Greenwood expressed dissatisfaction with the lack of information that had been sent to Members, particularly the failure to supply a calendar of dates (requested at UTC #62 in September). He felt that the Membership has been ignored. Adams is editor for the new edition. The schedule has been impacted by his workload at Stonehand. There was discussion of the tradeoffs: Do we want to delay the schedule and face the consequenses that might occur with Addison-Wesley? Do we do the 'right' thing by delaying it and getting it correct? Do we put down a new, more accurate, schedule? How we can make the schedule and still get people enought time to review the material? We can compress the time for review (maximum time requested was 8 weeks; actual time will be less). Members agreed that they can live with 4 weeks for review. It was agreed that pre-copy edit text will be distributed to Members for review at the same time that it is sent to AW. This is currently scheduled for March 17. Mark Davis said that if any dates slip, the review time will also, so that Members will not be denied review time. If there are real problems with the text, members will have time to get the changes back to the Consortium, and they ca n be merged with copy edit corrections from AW, to form the final camera-ready text. The calendar as distributed at the UTC has these dates: 3/19 -- Pre-copy edit text sent to AW, and shipped to Members as well. 4/1 -- Deadline for submission of CD-ROM material to Rick McGowan. 4/1 -- Lori Hoerth will make the mapping table s available to Members. 4/1 -- Code Charts done. 4/14 -- Master for making the CD-ROM "silver" to AW on April 14. 4/21 -- Deadline for review comments to Editor. 4/21 -- Editorial SC meeting if needed. Peter Edberg commented on problems with 1:many and many:1 mappings. Action Item #3 - Peter Edberg: Draft principles for mapping of characters. East Asian and 8859 series are the only tables that the UTC has to check. John Jenkins has been overseeing the work that Cora Chang has done to verify the East Asian tables. Vendor Tables are the responsibility of the particular vend or. Conformance Issue: ------------------ References: UTC64-95-006, 007, and 015. Discussion centered on the variant UTF forms, and their relationship to conformance. Aliprand moved a motion worded by Becker, seconded by Greenwood: That the UTC change the conformance clause to permit conformance of some other forms of use besides the 16-bit form of Unicode. After discussion, and at Becker's request, Aliprand withdrew the motion. Edberg moved, seconded by Cheng: That the UTC adopt UCS-2E as the basis for the conformance clause (and that the Editor adjust the existing conformance clause accordingly) and that Unicode 1.0 treatment still be conformant. Vote: Yes 9, No 3, Abstention 1, Absent 2 (Microsoft, AT&T). Davis moved, McGowan seconded: That Mark Davis is authorized to ensure that the UTF-8 algorithm operating on UCS-2E data provides exactly the same result as when it is applied to UCS-4 (ISO 10646) data. In discussion of the motion, it was noted that UTF-7 is defined as operating only on 16-bit values. Vote on UTF-8 consistency: Yes 12, No 0, Abstention 1. Motion passed. Davis moved, Honomichl seconded: In view of the additions and in recognition of changes to the conformance clause, that the Version in the book be Unicode 2.0. This motion was withdrawn. The book be called Unicode 2.0 in recognition of changes to the conformance clause. Vote: Yes 10, No 2, 1 Abstention. Motion passed. Honomichl requested that the reason for his opposition be recorded in the Minutes: he does not like the conformance clause. McGowan opposed the motion because there are new characters encoded. "Semantics" with respect to Conformance: ---------------------------------------- It was agreed by consensus that "semantics" with respect to conformance means character identity, and should not be construed to include character properties. Vietnamese Combining Tones: --------------------------- Reference: UTC64-95-009. Adams (proxy for Ecological Linguistics) moved: To deprecate the 2 Vietnamese combining tones and use the standard combining equivalents instead. Action approved by consensus. ZWJ Issue: ---------- Reference: UTC64-95-008. The use of the ZWJ and the ZWNJ is an issue in Devanagari, when used with a virama. Davis illustrated the various cases. Adams argued that the proposal in his discussion paper is closer to ISCII 91 (Indian national standard); that it provides easier lexocal recognition of the dead consonant; and that it enables easier encoding of an isolated dead consonant. Proposal approved by consensus. Combining Greek Koronis ----------------------- U+0343 Combining Greek koronis should be deprecated, as it was a mistake. Agreed by consensus. BiDi Issues: ------------ References: UTC64-95-011, 012. It is proposed to deal with the problem of BiDi reordering (as described in UTC document 011) by fixing the examples. Approved by consensus. The proposal on BiDi vertical rendering (as described in UTC document 012) was approved by consensus. Width Issue: ------------ Reference: UTC64-95-010 The following principle was agreed to by consensus: We should refer to width only where we have explicit (full and half-width) pairs. In any other case, width is implied. R-L Range: ---------- Adams proposed that the range reserved for R-L scripts be removed. Agreed by consensus. Apple, Taligent left at 6:15 pm. Davis gave the Chair a written statement on Taligent's position re combining/non-spacing terminology. Effect of ZWSP on Rendering: ---------------------------- It was agreed by consensus that the ZWSP should not affect rendering. The Editor was instructed to clarify the wording where it implied otherwise (e.g., in Devanagari shape forming). Combining vs. Non-Spacing: -------------------------- Davis: Use "non-spacing" for diacritical marks, etc., and use "combining" only for Devanagari. Adams: Use "combining" except in those cases where "non-spacing" vs. "spacing" is the subject at issue. Becker: "Combining" is relatively context-free, whereas "non-spacing" implies a "dead key" type of implementation. Adams (for Ecological Linguistics) moved, seconded by : Use "combining" except in those cases where "non-spacing" vs. "spacing" is the subject at issue. Vote: Yes 10, No 1 (Taligent proxy), Abstention 1 (Sybase). Motion passed. New Scripts: ------------ Adams (for Ecological Linguistics) moved, seconded by Ksar: Package Tibetan (with the forthcoming text), and Ethiopian as a UTR. Issue it for review, put it at the Web site (if feasible), and submit the UTR as a liaison contribution to WG2. Vote: Yes 8, No 1 (NeXT), Abstention 2. Motion passed. McGowan said that he voted against the motion because the Tibetan proposal included changes that he had never seen. Action Item #4 -- Glenn Adams The UTC strongly prefers that the BSI and the Unicode Consortium submit a joint proposal on Tibetan. Adams is charged with contacting the BSI (through Peter Lofting) to request delay of the BSI proposal until June, and to ensure that Lofting does not submit an independent proposal to WG2. Ksar moved, seconded by Greenwood: That the UTC recommend a U.S. co-editor for ISO/IEC 10646. Ksar nominated Glenn Adams for the position. Adams said that his acceptance of the position would be contingent upon travel funding by the Unicode Consortium. Aliprand pointed out that such travel had not been included in the Consortium's current budget, and would have to be discussed by the Officers. Vote: Yes 10, 1 No. Motion passed. Ksar moved to table the remaining agenda items. Agreed by consensus. It was further agreed that discussion on remaining editing issues would be continued by e-mail. Whistler moved to adjourn the meeting (6:50 pm).