ICU development team meeting
Time: 10am-11am PST, 1pm-2pm EST
Date: 2001-jan-10

Attendees:
    Daniel Chen (Lotus)
    Mark Davis (IBM)
    Steven Loomis (University of Malta)
    George Rhoten (IBM)
    Markus Scherer (IBM)
    Jim Snyder-Grant (Lotus)
    Syn Wee Quek (IBM)
    Raghuram Viswanadha (IBM)
    Vladimir Weinstein (IBM)

Minutes taker: Markus Scherer

Agenda items:

1. ICU 1.8 release plan (proposed by Markus and accepted)
  Release date 2001-mar-16
  Implementation complete 2001-mar-02
  API freeze 2001-feb-16
  (Same milestone definition as with ICU 1.7)

2. Windows codepage fixes
  - Issue: the ICU/IBM codepage data for Windows-12xx codepages is missing
    some newer mapping entries; Lotus is working on updating them
  - Jim: The data is updated, sent to NLTC (IBM codepage keepers),
    jitterbug submitted and ready to check in .ucm files
  - Mark: George has collected codepage data from Win2ksp1, should be compared to updated files
  - Markus: Please canonicalize .ucm files using ..../makeconv/misc/canonucm.c tool

3. George collecting codepage data in XML format
  - George: Code to convert validity tables ICU->XML,
    collect codepage data in better quality, generate new XML format according to latest UTR,
    going to publish on oss all ICU codepages, internally all collected
  - In the future, we will need validity tables conversion XML->ICU
  - Mark: Ram has almost completed a JNI wrapper for ICU conversion;
    about four times faster than Java conversion on large buffers

4. Collation reimplementation status
  - Vladimir: UCA tables in flat format, data structures defined,
    collation element retrieval & sort key generation implemented,
    working with Syn Wee on tailoring,
    working on passing UCA conformance tests, working on compare
  - Precanned UCA table will be included in ICU in some way, binary file some 75kB right now
  - Progressing towards completion for ICU 1.8
  - Jim: We have additional syntax to override default rules; true/false; true->don't use UCA data?
    Vladimir: possible, need to decide on whether we want to support this
    Mark: don't see need for this
  - Jim: Use of custom collation data?
    Vladimir: Put them into variant resource bundles, copy and modify parent bundle's collation rules
  - Mark: We will update the design doc soon
  - Jim: IBM Japan has started a discussion about redefining Japanese ICU collation order
    Mark: We want to use the same algorithm, need to discuss with all involved teams, may need variant data

5. Next meeting: Wednesday, 2001-jan-24, 10am PST/1pm EST