Re: Why people still want to encode precomposed letters

From: Mark Davis (mark@macchiato.com)
Date: Wed Nov 19 2008 - 11:51:01 CST

  • Next message: Erkki I. Kolehmainen: "VS: Why people still want to encode precomposed letters"

    Please
    don't read too much into my statements on this list. I am speaking in
    an official capacity as the president of the Unicode consortium only
    rarely, and I make it very clear when that is the case, such as in
    http://www.unicode.org/book/aboutbook.html#Foreword
    . So this, and other messages, are not ex cathedra!
    As to the matter at hand, there is not complete agreement on the goal of the
    named sequences. However, specifically, we aren't planning to add named
    sequences for every possible sequence that someone could reasonably want to
    appear with an optimal presentation* in fonts. For example, someone
    mentioned that it is appropriate to have an uppercase delta with a variety
    of combining marks for use in mathematics, but that is not appropriate for
    named sequences. So one should not be viewing named sequences as *the*
    source for font developers to see which sequences they should account for in
    their fonts.

    As far as the sequences used by languages, the Unicode locales project
    (CLDR) does have the capability to represent that, and does so for the
    languages it covers. It does not, of course, cover all languages, and will
    continue to be improved over time.

    http://www.unicode.org/cldr/data/charts/by_type/misc.exemplarCharacters.html

    * I say "with an optimal presentation". What users want is for the accent to
    appear in the right location. Whether this is done by mapping the sequence
    to a precomposed glyph, or by using anchor points, is an issue for the font
    designer.

    Mark

    On Wed, Nov 19, 2008 at 02:44, Andrew West <andrewcwest@gmail.com> wrote:

    > 2008/11/18 Kenneth Whistler <kenw@sybase.com>:
    > >
    > > Andrew West responded:
    > >
    > >> <http://www.unicode.org/Public/UNIDATA/NamedSequences.txt>
    > >
    > > To which my comment is assuredly not. Unicode named sequences
    > > are not nor have they ever been intended to serve as
    > > guidance for font developers about what glyphs should or should
    > > not be supported for fonts.
    >
    > UAX 34 (which you wrote) would seem support your confident assertion
    > that named sequences are not intended to act as guidance for font
    > developers. But I wonder how widely accepted this orthodoxy is amongst
    > UTC members.
    >
    > For example, on Saturday 10th September 2005 Mark Davis wrote on the
    > Unicode list <http://unicode.org/mail-arch/unicode-ml/y2005-m09/0190.html>
    > :
    >
    > "I think we are in agreement on named sequences; they should give
    > guidance to font developers as to which char sequences may need a
    > precomposed glyph."
    >
    > Perhaps in the intervening three years his understanding of named
    > sequences has changed, but this statement on the public Unicode list
    > by the president of the Unicode Consotium has certainly informed my
    > understanding of what named sequences are about.
    >
    > Andrew
    >
    >



    This archive was generated by hypermail 2.1.5 : Wed Nov 19 2008 - 11:56:34 CST