From: Mark Davis (mark.davis@icu-project.org)
Date: Tue Dec 26 2006 - 20:32:34 CST
Having a precomposed character isn't really the issue. In OpenType, a font
vendor can supply a glyph for a non-composed character essentially just as
easily as supplying a glyph for a precomposed character, by adding a
ligature or using mark positioning. The more interesting issue is for the
font vendor to know which glyphs (whether for precomposed characters or not)
are needed for the markets they are targetting. It's typically not all
possible glyphs, since that is simply too expensive to do in all faces.
Mark
On 12/26/06, Michael Everson <everson@evertype.com> wrote:
>
> At 17:27 -0700 2006-12-26, John H. Jenkins wrote:
>
> >This was done some fifteen years ago. Now, all the major
> >Unicode-supporting platforms do the right thing and this kind of
> >hack is no longer needed.
>
> What doesn't work is fonts presenting users with decent glyph
> representation for accented Latin characters which are not
> precomposed in the standard.
>
> >Moreover, adding precomposed accented forms to the standard at this
> >point would create dangerous instabilities for processes such as
> >normalizing and will not be done. There are no longer any technical
> >needs to do it and strong technical objections to it being done.
>
> Support for disadvantaged languages is still suboptimal at best.
>
> I hate to say it, but it is true.
> --
> Michael Everson * http://www.evertype.com
>
>
-- Mark
This archive was generated by hypermail 2.1.5 : Tue Dec 26 2006 - 20:34:32 CST