🚀 go-pugleaf

>What is the min, max, and median number of radicals per glyph?  How about
>per word?  As was pointed out by someone else, Japanese may take multiple
>glyphs to encode a word.

Looking at Han characters in a probabilistic sense probably is not going
to help much, since the positioning of radicals varies widely between
characters.

The majority of rare characters needed for normal purposes can be composed
by putting a "productive" radical on the left, top, bottom, or right side
of a character which already has an existing codepoint.  There are two
major catches:  (1) some rare characters cannot be expressed in this manner,
and (2) allowing the display of arbitrary characters using this sort of
composition does not mean that their components will be aesthetically
spaced.

>In terms of benefit to the Japanese, any answer over and above "2" is more
>expensive than raw encoding for a 16 bit font, or "4" if a 32 bit font is
>used.

A 16 bit font is insufficient for encoding rare characters, whichever way
you look at it, although having 16-bit CJK unification and a user-defined
character facility may be sufficient for an average user.

-----
John B. Melby
Fujitsu Limited, Yokohama
melby@yk.fujitsu.co.jp