Celeste encoding (was: Duplicate messages in Celeste)

Yoshiki Ohshima ohshima at is.titech.ac.jp
Wed Mar 15 21:18:31 UTC 2000


  Hi Henrik,

  You're mentioning Zaurus so I'm supposed to respond:-)

> There's a free, high-quality Bitstream Cyberbits font with a full Unicode
> char set. It's almost 13MB big! I guess our Japanese squeakers would
> appreciate it, for example.

  Umm, yes and no.  Note that there is no (will no) single
Unicode font that fulfils Chinese people, Korean people,
*and* Japanese people's need.  A Unicode font merely fulfils
Chinese, Korean, *or* Japanese.  Each "country" have to have
their own glyph based on the same encoding.  This concept
doesn't match well with Squeak's 'dot-by-dot identity'.

> Just write a method or two in Squeak for mapping the characters (I can show
> what methods to write but don't know Unicode or character encodings in
> general) and you could put all these characters in a bitmap font and use
> this like in any regular Squeak font for text. With the fancy anti-aliasing
> and all, I can imagine that the Japanese characters will look really sharp
> on eg. the Zaurus' color LCD.

  Yeah, anti-aliased fonts on Zaurus screen would very nice.
Note there are many TT fonts based on JIS X 0208 character
set.

> Unfortunately, a 13MB Windows font is a bit more than my
> Mac font software & RAM can swallow so I haven't been able
> to open it up and look inside it.

  Yeah, we need some font cache mechanism that John Duncan
suggested before.

  I don't try to be pessimistic, and want to have cool font
handling mechanism in Squeak.  I just want to clarify that
"Unicode" is not universal answer for multilingualization.

  -- Yoshiki





More information about the Squeak-dev mailing list