🚀 go-pugleaf

In article <1ippgmINN7af@life.ai.mit.edu>
	glenn@wheat-chex.ai.mit.edu (Glenn A. Adams) writes:

>>Different Characters with different stroke counts are mapped to a
>>single code point in 10646/Unicode.
>
>For those in the listening audience who don't read Hanzi or Kanji, I should
>say that Ohta-san is correct, but in a completely frivolous way.  Stroke
>counts are a bit subjective.  Some people count some elements as distinct
>strokes, whereas others do not.

Wrong. That's not what I'm referring.

>However, there is no systematic difference
>in the way that Japanese count strokes versus the way that Chinese count
>strokes versus the way that Koreans count strokes.  The ways that CJK
>standards bodies count strokes are extremely uniform;

True, but it is misleading.

Consider, for example, a very common radical used for plant-related
characters, such as grass and flowers (character code from 827C to 864F
of DIS 10646-1.2).

Its shape in Japanese or mainland China is

        *        *
     ****************
        *        *

whose stroke count is 3, while its shape in Taiwan and Korea is

        *        *
     *******  *******
        *        *

whose stroke count is 4. But, these different characters with differently
shaped radicals are unified in Unicode.

>therefore, the
>unification of CJK characters into single Han code elements does not 
>remove discriminating information;

Therefore, the stroke count information, which is "THE EXTREMELY UNIFORM"
property of characters, are destroyed by Han unificaiton.

This is the reality.

>This is a bit like saying "Japanese (or English or another language) is
>context dependent.  Don't use it."  Come on Ohta-san, return to reality.
>Try making a rational argument.

I have come again. I'll be happy if you stop dreaming and shut up.

						Masataka Ohta