Oh heck, I can tell you what's wrong with it.  Han unification for one.  The unicode consortium adopted the mantra "characters not glyphs".  So they coded up all the Hangul characters and unified characters that are called the same and mean the same thing across some asian languages.  The problem is that they do not draw the characters the same.  This makes it awkward to mix certain asian languages because they are "sharing" characters and there isn't enough contextual information to tell if you should draw them like they do in China or Japan.  

Additionally, ASCII has been encoded into unicode twice - specifically to support half and full width typography (in violation of above mantra) for Japanese.
 
Unicode wasn't a bad first attempt, but it ought not to be the last one.


On Wednesday, March 23, 2005, at 11:51AM, David Farber <dfarber at numenor.com> wrote:

>
>Yoshiki - I am intrigued by your statement.  Could you say more about what you think Unicode's shortcomings are?  And what a good multilingualization standard might look like?
>
>Thanks,
>david
>
>At 10:34 AM 3/23/2005 -0800, you wrote:
>>  But, if you replace the word "multilingualization" with
>>"object-oriented programming", and "Unicode" with "Java", you might
>>get some of the idea ^^;
>>
>>-- Yoshiki
>>
>>
>--
>David Farber
>dfarber at numenor.com
>
>
>