On Fri, 15 May 1998, Rick McGowan wrote:
> The "Unicode CMAP" is just a convenient crutch, isn't it? Fonts could
> implement glyph sets for Hieroglyphics or Linear B or whatever and not have a
> "Unicode CMAP". It should not matter what the encoding for surrogate codes
> is -- that's irrelevant as far as the FONT is concerned.
I agree with you on this.
> I think the "Unicode CMAP" is convenient for some things to some people.
> Rendering of Unicode text for display is a complex problem. To do it, you
> need one or more ordered sets of glyphs, and a means of translating from your
> stream of character codes into a corresponding stream of glyph codes. Just
> having a "Unicode CMAP" doesn't get you the second piece. It will get you a
> long way in the Latin script and in CJK, but it won't get you much past that
> into the more complex scripts.
I guess you meant various Indic scripts and Thai, Arabian and Hebrew
scripts by more complex scripts. As a matter of fact, Korean Hangul
script has a lot of things common with those scripts and needs to be
treated similarly if one wants to support it to the fullest extent(There
are well over 50k Hangul syllables to be composed out of Hangul
Conjoining Jamos in U1100-U11FF block) and/or in the most efficient(in
terms of memory and disk space required of storing and handling fonts)
way. More or less the same thing can be said of Chinese Hanzi(if one
wants to render over 50k of them). For all these scripts, something
different than the present approach which assums there's one-to-one
correspondence between code points and glyph indices is necessary.
Jungshik Shin
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:40 EDT