16 IDCs are sufficient?

From: mpsuzuki@hiroshima-u.ac.jp
Date: Wed Jun 10 2009 - 02:31:23 CDT

  • Next message: Hans Aberg: "Re: unicode to postscript"

    Dear Andrew,

    I found your proposal to SC2/WG2 N3643 to add 4 IDCs.

    In my understanding, you propose them as the tools for
    standardization works to classify the ideographic
    characters and evaluate the similarities of their shapes.

    I think it's good idea and I agree with the requirement
    of new IDC (a few proposed combinations are impossible
    to be described by existing IDCs), but I'm not sure if 4
    IDCs are sufficient for future works.

    You mentioned several scripts - although I'm unfamiliar
    with their detailed calligraphic/typographic rules, I
    heard that a composite glyph of Khitan small script can
    include 2-7 components. If the experts want to describe
    such composite glyph by single IDC and multiple
    components, they may request more IDCs. If such
    possibility is not unrealistic, I think it's a time to
    propose another larger block for extended IDCs.

    In fact, the addition of 4 IDCs will fullfill existing IDC
    block, so they should be carefully designed. The proposed
    IDCs are sampled from the IDCs for Old Yi (and you've shown
    they are useful for Nushu), not for CJK Ideographs. It can
    be arguable whether it is good choice to fill IDC block by
    the IDCs for Old Yi, not for CJK Ideographs.

    Among the mentioned scripts in your proposal, all of them
    would be coded out of BMP, so I wish coding new IDC(s) out
    of BMP won't be serious difficulty.

    # There can be an objection - if the standardization work
    # of completely unencoded script usually starts with font
    # using BMP PUA codepoints, it can be troublesome for some
    # existing systems to mix the components on BMP PUA and
    # new IDCs on non-BMP area.

    Andrew, how do you think about the possibility about
    the requirement of extra IDCs after your proposal, and
    the requirement of the IDC restriction to BMP?


    This archive was generated by hypermail 2.1.5 : Wed Jun 10 2009 - 02:38:20 CDT