ISO/IEC JTC1 SC22/WG20 N874
MINUTES - Malvern
ISO/IEC JTC 1 SC22/WG20
Meeting #21 - Internationalization
October 3, 2001
DATE: October 1-3, 2001
TIME: 9:00 a.m. to 5:00 p.m
LOCATION: Unisys
2476 Swedesford Road
Malvern, PA 19355
CONTACT: Arnold F. Winkler
Tel: 610-648-2055
Web site: http://wwwold.dkuug.dk/jtc1/sc22/wg20/docs/documents
Hotel information and driving instructions
Map between Philadelphia Airport and Unisys
Arnold explains the logistics.
Arnold announces, that he will not be able to apply again as convenor, when his term is over.
Some people (John Clews, Sandra O’Donnell) would like to call in for specific subjects – we need to define time for these and communicate to them.
875 |
Participants at the WG20 meeting in Malvern, October 1-3, 2001 |
Winkler |
2001-10-03 |
Participating at this meeting:
Alain LaBonté Canada
Kyongsok KIM Korea (South)
Keld Simonsen Norway
Ken Whistler USA
Arnold Winkler convenor
Arnold will act as secretary.
Drafting committee: Keld, Kim
Minutes Tübingen, May 8-10, 2001 |
Arnold Winkler |
2001-05-11 |
The minutes were amended:
From N813: Discussion about the merits of either ordering. Decision that the order stays as in the table which is the Fuþark order. Remove “…. which is the Fuþark order.”
Typo: AIX in agenda item 15, comments Baldev.
Minutes are approved as amended.
2002-05-20/22, Norway
2002, October, possibly Japan; AI Arnold to check with Maki. Possible backup Korea.
May 2003 – Canada.
882 |
Proposed disposition of comments for ballot for PDAM #1 for 14651 |
Alain LaBonté |
2001-09-30 |
883 |
Proposed disposition of comments to DTR 14652 ballot comments |
Keld Simonsen |
2001-10-02 |
872 |
Agenda for the WG20 meeting in Malvern, |
Winkler |
2001-10-01 |
Approved.
Resolutions from the SC22 plenary in Hawaii |
SC22 N3316 |
2001-09-26 |
01-30 no action necessary for WG20
01-18 Keld volunteers for liaison to IETF
01-22 Ballot on the update is SC22 level, WG20 will also need an editor for TR 10176.
01-23 10176 on ITTF web site: same as 01-22
01-24 Future of WG20 - wait for results, no WG20 recommendation agreed upon
01-25 DTR 14652: Keld will provide a proposed disposition of comments on October 2, 2001. This will include the SC22 proposed re-packaging. There is significant disagreement in what is contentious and what is not. Kim asks about the relationship between 14652 and 15897? Arnold (and others) explain the history of this.
01-26 15435 API: wait for SC22 letter ballot result. If the project is canceled, a TR with a description of the functionality can be proposed.
01-27 secretariat action.
01-28 no input required
01-33 Based on a presentation of Ksar at the SC22 meeting. No recommendation from WG20. SC22 needs to take into account, what Java has done for programming languages, as well as scripting languages, etc… Fundamental disagreements must be eliminated to accept 10646 as the generally acceptable character set. Ken explains a layered approach, Keld goes along with that. The committees with general language tools are the ones that are interested, not even COBOL. C and C++ with lots of application development need to take care of 10646, and I18N, and related subjects.
List of liaison and co-operations to and from SC22/WG20 |
Winkler |
2000-05 |
Using TR 14652 in COBOL 200x |
Ann Bennett |
2001-09-17 |
This is for information only.
No report
Keld: N830 was forwarded and created lots of e-mail discussion. Will be further discussed in C and C++ meetings. Plan for an ad-hoc on character sets in the SC22 plenary next year.
New POSIX standard in ballot.
TR 14766 – guidelines on national profiles and locales. Resolving PDTR comments.
Keld: N830 was forwarded and created lots of e-mail discussion. Will be further discussed in C and C++ meetings. Plan for an ad-hoc on character sets in the SC22 plenary next year.
No report, nothing to talk about. Arnold has asked Kimura-san to put the liaisons on the agendas for the upcoming SC2 and SC2/WG2 meetings.
No report, decide to drop the liaison. Resolution.
Alain: keyboard standards are being updated. New TR project proposal for evaluation of I18N in software.
Keld reports about the Price WaterHouse Cooper report – decision on recommendations is outstanding. Steering group is being formed between CEN and LISA and others.
Euro-locale workshop created – submitted from SC22 questions if CEN TC304 is an authorized body to register the locales.
Will be discussed, when John Clews is connected Tuesday 9:30.
TC37 met in Toronto in August. See N866. TR in preparation for extended language codes. The task force has no meetings scheduled, only e-mail discussion.
UTC moved from Baltimore to Mountain View. Technical items are covered in ballot comments from L2. Controversial issue : UTF-8s (now being worked as a proposed draft TR for Unicode). Unicode 3.2 will be in beta status shortly after the Singapore meeting of SC2/WG2, when the comments to PDAM#1 of 10646 are disposed. Next book 2003.
See SC22 resolutions 01-16, 01-17, 01-18, 01-18a, and 01-28 in N868. No action.
Action item list |
Winkler |
2001-05 |
Action items were reviewed, the SD-5 updated.
See SC22 resolutions 01-22 and 01-23 in N868
A letter ballot for approving the update of TR 10176 was authorized.
The suggestion in N838 was forwarded to ITTF as SC22 N3249, it was not discussed in SC22. Ken Whistler recommends that the table to be referenced be the Unicode character data base.
Keld would like to put it into the cultural registry. Ken speaking for the US indicates that the US would be opposed to that – no versioning and synchronization with other data tables is provided, it also does not fall into the “locales” that people would expect in a cultural registry. The name “registry” is detrimental to the acceptance by the USA, everybody could register whatever they want. Ken points out that the US position on comments to 10176, 14652, etc. are unanimous.
Kim: discussion about what characters might be used in some languages, but not in others. Examples are hyphen or underscore, which are allowed in some languages, but not part of the table; and e.g. combining marks that are allowed in Java, but are not part of the 10176 table either.
Draft 14651 tailoring delta for D P R of Korea |
Kent Karlsson |
2000-10-05 |
|
Proposed ISO/IEC 14651 CTT tailoring delta for D P R of Korea [re-order table in text form] |
CSK |
2001-09-03 |
|
Comments regarding N867, the DPRK proposed delta file for IS 14651 |
KIM, Kyongsok |
2001-09-27 |
|
Comments on PDAM #1 for ISO/IEC 14651 |
Kent Karlsson |
2001-09-27 |
|
Draft ISO/IEC 14651:2001 CTT tailoring delta for D.P.R. Korea – reorder table in plain text format |
Kent Karlsson |
2001-09-18 |
|
Summary of voting for the registration and PDAM#1 consideration of ISO/IEC 14651 |
SC22 N3318 |
2001-09-28 |
|
882 |
Proposed disposition of comments for ballot for PDAM #1 for 14651 |
Alain LaBonté |
2001-09-30 |
Alain presents his proposed disposition of comments as documented in N882.
Ireland: wants the Runes ordered differently – FuÞark. US can go along with that. Ken has to change the relevant section of the table.
Japan: placement of the table on the ITTF site.
JP-1: Reference to the table must be part of the amendment text – this links the repertoire covered in the standard at hand with the corresponding table. When we do new additions, the explanation of the previous tables needs to be in the addition.
JP-2: Japan is worried about the integrity of the table on the ITTF web site. Once the amendment is final, the URL will be fixed, and the table will be secured by ITTF. Backup copies by ITTF and the editor ensure constant contents, once approved. Neither contents nor name of the table will change.
JP-3: How to normatively reference electronically published standards.
USA:
US-1: repertoire of
Unicode 3.1 will be provided by Ken Whistler to the editor.
AI – Ken: Provide updated table
to the editor. The repertoire needs to
be Unicode 3.1.
US-2: accepted.
US-3: accepted
US-4: accepted, overlap problems in edge cases
Sweden:
SE-1: accepted. Informative note.
SE-2: accepted.
SC-3: accepted, text will be changed slightly.
SE-4: accepted.
SE-5: Ken explains the current order
Today EXT A (3400)
URO (21000+)
12 CJK compatibility characters
Now addition of EXT-B – Kent suggests a re-ordering, including additional 500 CJK compatibility characters. Accepted.
SE-6: rejected, making synchronization with Unicode Collation Algorithm difficult.
SE-7: not accepted, MIN & MAX are sufficient.
SE-8: deferred to the next edition
SE-9: not accepted at this moment, do with the next full edition of the standard
SE-10: accepted
SE-11: not accepted, not relevant for the amendment.
SE-12: Annex E – not provided by Sweden.
Full agreement on the discussed disposition of comments. Resolution about the progression of 14651, amd.#1.
AI - Ken promises the next table by November 23 latest.
FPDAM #1 sent to SC22 by end of November 2001.
---------------------------------
John Clews dial in (+44-1423-888-432):
ICU collation update will cover Unicode 3.1
Runes: any order is OK. Possibly preferred the transliteration order.
Burmese, Khmer, Mongolia: asking if there is any additional input – NO ordering except the basic character ordering has been suggested.
------------------------------------
Discussion of Korean ordering (Kim):
Prof. Kim presents his paper N876, comments to the DPRK proposed delta and Kent Karlsson’s proposal. There are different ordering rules in the two Koreas. Explanation of transliteration of Korean with varying methods for consonants.
DPRK might propose 8 Old Hangul letters.
It would be nice to have common ordering rules between both Koreas. This needs to be agreed upon by the two Koreas, no need to get into that discussion. SK does not have an order for the Old Hangul. SK will publish 14651 with a Modern Hangul order specification.
Normalization:
Ordering works fine for complete syllables. 14651 recommends normalization before ordering, if independent letters are to be sorted. (UCA does require normalization). If only compatibility characters are ordered, the order is OK, problems will only come up when old and new code pages are mixed.
See SC22 resolutions 01-25 in N868
CEN CWA – European Culturally Specific ICT Requirements |
CEN CWA 14094, ESR:2000 |
2000-10-31 |
|
Rules for the use of IJ in Netherlands pubic records |
Johan van Wingen |
2001-09-13 |
|
864 |
CEN prEN 14142-1: Components of postal addresses (paper copy only – request) |
CEN/TC331 |
2001-05-05 |
Using TR 14652 in COBOL 200x |
Ann Bennett |
2001-09-17 |
Discussion about the meaning of the word “contentions”. Some sections are controversial and have been so for many ballots – the dispositions of comments from earlier ballots are non satisfying.
N883 was produced by the editor as the preliminary disposition of comments (50 pages). The document was distributed as input to the discussion.
Sandra O’Donnell used teleconferencing to contribute to the discussion.
The SC22 plenary has passed a resolution how to re-package the DTR with markings, which of the sections are controversial. Any resulting document will be balloted again as a DTR according to SC22 resolution 01-25.
Sandra: 4 major objectionable areas: monetary with dual currency, time zone, repertoire map, and transliteration.
Objection #16: the question is, if e.g. French Francs and Euros are one currency ? The enhancement is controversial, the whole section.
Keld: the US proposal to use alt_currency_symbol is not acceptable to the editor.
Section 4.5 is marked as controversial.
Objection #32: transliteration, section 4.9.
The method proposed here is not generally usable, some languages (Hebrew, Japanese, etc…) can not be transliterated. Sandra believes that transliteration does not belong into the TR. Locale concept extensions are not workable – example were ICONs.
Keld says, that transliteration only works into the locale from various languages. All rules would need to be defined in one locale – it does not scale well.
Section 4.9 is controversial.
Objection #30: Time zones, section 4.7, the extensions beyond POSIX 2 are controversial. Only works for 1 time zone. Is not culturally specific. There are more controversial parts than the time zones.
Section 4.7 is controversial.
Objection #53: – repertoiremap, section 6
Section 6 REPERTOIREMAP is controversial.
Section 7 – conformance is controversial (e.g. Technical #69).
Keld: TR type 1 are failed standards and thus do have conformance sections.
Sandra promises to send to Keld wording about the combining classes.
Long discussion about the usefulness of TR 11017 and other work of WG20.
Keld leads through the document N883.
Finland:
Keld will ask the Finnish NB, if their comment about structural issues can be used .
French:
1. Audience: French comments are accepted, their text will be used.
2. France asks for assertion of POSIX compatibility. Keld will add a line to reference the new POSIX standard “this TR is aligned with POSIX 200x“.
3. multiple currencies are controversial. Some NBs want it, some don’t.
4. No change to the draft.
German:
LC_CTYPE classes: confusing specification, can not be changed due to existing implementation.
Monetary section is deemed controversial. The German comments are rejected. Statement is needed that applications conforming to the new standard can also process old locales. Old applications will not run new formats.
LC_XLITERATE is controversial, text will not be changed. Used in some LINUX implementations.
Sweden:
Answer should be along the lines of the answer to Finland. Controversial sections are marked, invite Sweden to contribute to Annex D. Point to POSIX practice for need of repertoiremap in string representation.
Ken indicates the US position to ensure, that all controversial sections are clearly identified, and that the DoC shows how the TR will be changed.
SE-1: not accepted.
SE-2: no change to text, rejected.
SE-3: rejected, no change to the text, existing POSIX practice
SE-4: string format – rejected, no change
SE-5: not accepted, no change to the structure of the FDCC set
SE-6: not accepted
SE-7: not accepted – transliterate section is marked as controversial
United Kingdom:
The comments are rejected. Sections are marked controversial, the UK is invited to submit dissenting opinions to Annex D.
Keld indicates that he will re-write parts – so the comments are partially accepted.
United States:
The numbers refer to the USA objections with the same numbers.
Keld agreed to remove argumentative sentences in the disposition of comments and to concentrate on specifying, which comments are accepted and which are not. The disposition will also identify, which objections are not accepted because of POSIX.
US-1: not accepted
US-2: not accepted
US-3: accepted
US-4: use of ranges – Japan had asked for that in an earlier comment. This function requires more complex parsers, the tables are less readable. Accepted, text will be removed, tables modified accordingly.
US-5: rejected
US-6: accepted
US-7: accepted
US-8: rejected, add “decimal” before digits
US-9: accepted
US-10: not accepted
US-11: same as German comment, asking for removal of combining classes from the text (and a definition of class). LC_CTYPE extension to “class” and “map” keywords are controversial.
US-12: rejected, keyword “width” is marked as controversial
US-13: “map” keyword is controversial partially accepted.
US-14: not accepted
US-15: add fullwidth digits to repertoiremap. Spaces handling will not be changed because it is POSIX practice (class definition is not sufficient). Partially accepted, some changes to the text will be made.
US-16: The complete section LC_MONETARY is marked as controversial. Comments are not accepted.
US-17: rejected, words come from POSIX
US-18: not accepted, section marked as controversial
US-19: partially accepted, some rewording for clarification will be done.
US-20: marked as controversial
US-21: rejected
US-22: not accepted, complete section marked as controversial
US-23: Not accepted, complete section marked as controversial
US-24: see above
US-25: see above
US-26: not accepted
US-27: accepted: note that various formats of ISO 8601 exist will be added at the beginning of the LC_TIME zone.
US-28: not accepted
US-29: not accepted
US-30: editorial correction accepted, text change not accepted.
US-31: not accepted
US-32: LC_XLITERATE: not accepted. The complete section is marked as controversial. Clarification for target language will be added.
US-33: accepted (consistency)
US-34: accepted. String. Clarify format.
US-35: accepted, clarification will be added.
US-36: accepted, clarify intent.
US-37: not accepted
US-38: profession is not part of the name. Clarify.
US-39: editorial comment accepted, comment not accepted.
US-40: accepted
US-41: accepted
US-42: define CEPT mailcode and the fact that this is simply a string.
US-43: accepted, clarifying text will be added
US-44: This does not work well for the USA. E.g. Addressee is not part of the LC_ADDRESS . It is necessary that a note be added to explain that fact. Additionally, an alternative format (U+xxxx) need to be added to make it understandable. Examples for USA and Korea will be added.
US-45: accepted, clarifying text will be added
US-46: accepted,. string field for extension needs to be added
US-47: accepted
US-48: accepted
US-49: accepted
US-50: accepted
US-51: accepted
US-52: not accepted.
US-53: REPERTOIREMAP – the complete section is marked as controversial. Mainly rejected.
US-54: noted.
US-55: accepted
US-56: accepted
US-57: accepted
US-58: accepted
US-59: accepted
US-60: accepted
US-61: accepted
US-62: accepted in principle, make consistent with 14651
US-63: accepted
US-64: accepted
US-65: accepted
US-66: accepted
US-67: accepted
US-68: partially accepted, will be made consistent. Charmap can contain other entities, such as anchors for re-ordering in formats that are not Unicode identifiers.
US-69: not accepted. Section 7 is marked controversial.
US-70: accepted
US-71: rejected, this is how it is used in POSIX
US-72: accepted. Solution will be found (and reviewed).
US-73: accepted.
US-74: accepted, will be fixed.
US-75: accepted.
Progression:
Editing group: Keld, Alain, Kim, Ken, Sandra, Arnold
Editor |
New disposition of comments N883R |
October 5, 2001 |
Ed. Group |
Review by the editing group (approval by default) |
October 10, 2001 |
NBs |
Annex D input to the editor |
November 21, 2001 |
Editor |
New DTR document |
End of November |
Ed. Group |
Review of the new DTR draft |
December 11, 2001 |
Editor |
Final DTR document for ballot |
December 14, 2001 |
|
Submission for DTR ballot |
December 17, 2001 |
See SC22 resolutions 01-27 in N868
PDTR 14766 Guidelines for POSIX National profiles and national locales |
SC22 N3234 |
2001-05-24 |
|
Text for registration and CD ballot of ISO/IEC 15897 Registration of cultural elements (replaces N845) |
Keld Simonsen |
2001-07-01 |
|
Proposed ISO/IEC 15897 registration: |
TIEKE |
1999-01-25 |
|
Proposed ISO/IEC 15897 registration: |
Norwegian Tech. Ctr. SC22 N3264 |
1998-12-22 |
|
Proposed ISO/IEC 15897 registration: |
ITS |
1998-12-22 |
15897 is in CD ballot now.
See SC22 resolutions 01-26 in N868. No new document. Waiting for the SC22 letter ballot result. Project could be withdrawn and a TR for the functionality proposed.
See SC22 resolutions 01-33 in N868. Preparation for a character set ad-hoc at the next SC22 is needed.
Resolutions from CLAUI meeting in France |
CLAUI meeting |
2000-10-19 |
|
Notice and call for documents for a CLAUI meeting in Malvern, PA October 4-5, 2001 |
Arnold Winkler |
2001-06-08 |
|
Cancellation notice for the CLAUI meeting in Malvern, October 2001 |
Arnold Winkler |
2001-08-16 |
Arnold explains his reason for canceling the meeting (no contributions, no agenda item proposals, economic situation).
Resolution: for the next CLAUI meeting WG20 nominates Keld and Alain as representatives.
Rules for the use of IJ in Netherlands pubic records |
Johan van Wingen |
2001-09-13 |
|
Report on character set policy in the Netherlands |
Johan van Wingen |
2001-09-21 |
See SC22 resolutions 01-30 in N868
2001 Systematic review of International Standards |
SC22 N3222 |
2001-03-14 |
No action from WG20 is required.
Progression of IS 14651 Amd.#1 and TR 14652 are in their respective agenda items.
Action item list |
Winkler |
2001-10-03 |
For new action items see the updated SD-5.
The resolutions were reviewed and approved.
The meeting was adjourned on October 3, 2:45 pm.