Mac OSX + 17,500 Kanji

From: Stephen Cremin (
Date: Fri Mar 16 2001 - 21:50:38 EST

I didn't catch Steve Job's keynote in Tokyo but I believe he announced that
Mac OSX (out 24 March) would be supplied with Japanese fonts representing
17,500 kanji. Presumably, Mac OSX is using Unicode to represent text

I database information in Japanese (as well as Chinese and Korean) and I
generally takes the lowest common denominator approach. I approximate
characters to what can be displayed back to me [in the Japanese version of
Mac OS8.0 running various language kits], with a note of the correct Unicode
codepoint for future reference. Another factor that dictates how accurately
I store kanji correctly is what I can expect to present to people over the

Presumably using UTF8 encoding, other Mac OSX users who install the Japanese
fonts can now display these 17,500 kanji leading to greater accuracy in
online information. But, of course, Mac OS X will still be a minority even
among Mac users for the next year or more. Forgive me for being so insular,
but what is the situation in the non-Mac world? How many characters can the
typical Japanese-enabled UNIX or IBM compatible handle?

I understand that different fonts may not render all codepoints, but what is
typical for a pre-installed Japanese user's system in the non Mac-world?
And what changes are taking place in the near future? I don't want a Mac vs
IBM vs UNIX debate, just an idea that as a web developer, how many kanji can
I presume to work with and in what timeframe?

And any announcements from Apple on Chinese and Korean fonts? If this
17,500 kanji refers to a specific "Unicode font" then I presume there are
various Chinese, Japanese and Korean flavours. And what other advances are
there in Mac OSX in terms of tagging unicode to aid sorting, etc? It would
be more than wonderful if I could assume that every "carbonised" application
will allow me to sort in my choice of language and its variants [North
Korean sort order, South Korean sort order, etc], search two-byte text
reliably, searching two-byte text "intelligently" (allowing for semantically
correct variants, allowing for commonly mistaken kanji, etc), etc? But I'm
not so optimistic.

Stephen Cremin

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:20 EDT