I have a question about using the PUA (Private Use Area) in TrueType
fonts with a Unicode cmap subtable on Windows and other platforms.
Mozilla is currently implementing MathML. The math experts are
attempting to get some math symbols into Unicode, but there seems to be
a concern that not all of their symbols will be accepted, for whatever
reason (they aren't "characters", they're already encoded, etc).
So we are wondering whether it would be wise to use Unicode-encoded
TrueType fonts for math, and to use the PUA for those symbols that end
up being rejected by the Unicode committee.
Specifically, do the Windows GetTextExtentPoint32 and ExtTextOut
functions do anything special with the PUA codes on any of the localized
versions of Windows (e.g. US, Japanese, Arabic, etc)? E.g. do the bidi
versions do some "helpful" (automatic) reordering of characters in the
PUA range? Or anything else that might affect the results of those APIs?
A more general question would be whether it is wise to use the Windows
Unicode cmap subtable if we want to use the TrueType font on other
platforms like the Mac, etc. Would we be better off sticking to the old
256-element fonts like the Windows "symbol" encoding? Would it then be
easier to come up with a cmap subtable that actually works on the Mac
and other platforms (e.g. OS/2)?
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:56 EDT