L2/01-031

Report from the IRG meeting in Seoul, December 4-7, 2000


The sixteenth meeting of the Ideographic Rapporteur Group was held in
Seoul, Korea from 4 to 7 December 2000.

The L2/Unicode delegation consisted of Hideki Hiura (Sun), John H. Jenkins
(Apple, head of the Unicode delegation), Michael Kung (Microsoft), and
Michael Yau (Oracle).

The purpose of the meeting was to respond to requests to extend the mapping
data issued by the IRG to include contributions from the Democratic
People's Republic of Korea (North Korea), and to respond to errata reported
in the glyph data in the current proposal for Vertical Extension B.

Unicode had objectives in both cases.  With regards to the DPRK mappings,
the IRG had been mandated to incorporate those mappings if possible without
disrupting the existing data (that is,without changing unifications). With
regards to the errata in Vertical Extension B, the fundamental goal was to
make sure that Vertical Extension B and the publication of ISO/IEC 10646-2
and Unicode 3.1 stayed on schedule.

The DPRK proposed mappings had been sent to the IRG without covering
materials. In particular, no copies of the relevant North Korean standards
were made available to the IRG. Examination of the mapping data lead to the
suspicion that different unification rules than those adopted by the IRG
had been used in generating the data. In any event, without additional
materials which would allow the IRG to proof the work done, it was clearly
impossible for the mappings as suggested to be adopted.  The IRG accepted
in principle the idea of including DPRK mappings into its data but referred
the matter back to North Korea for additional information.

The bulk of the work of the meeting was spent proofing glyphs in Vertical
Extension B.  It was divided into sections and all IRG attendees were
assigned a section to proof with the exception of the Rapporteur and the
one gaijin in attendance. The results were correlated and discussed by the
entire body. Care was taken to make sure that glyphic changes would not
disrupt unifications. For example, where a character from CNS 11643-1992
was incorrectly identified with a character in the KangXi dictionary, that
identification was allowed to stand as changing it would force a
deunification and change the mappings in Vertical Extension B.

With the errata correlated and accepted, the IRG adopted a schedule for
further work. The US/Unicode objected to the schedule as it would probably
not leave enough time for a suitable font to be provided to Unicode for
printing Unicode 3.1 on schedule.

The IRG also agreed to supply radical-stroke data to Unicode for use in
printing a radical-stroke index to the full ideographic repertoire of
Unicode 3.1, with the proviso that Unicode understand that the data are
insufficiently proofed.  A schedule for initial work on Vertical Extension
C was also adopted as were standards for data required in Vertical
Extension C proposals.