Re: [tlug] keitai encodings

[Shimpei Yamashita <shimpei@example.com>]

On Fri, Jul 11, 2003 at 08:22:39AM +0200,
David Oftedal wrote:
> >
> Ach. How hard can it be to implement customizable input methods (by way 
> of Java, for instance) and unicode in mobile phones... I just hate how 
> mobile phones are incompatible at so many levels across borders.

1) Decent input methods and Unicode fonts cost money--licenses,
development/testing, memory chips--and occupy valuable space within the cell
phone that could be better used by, for example, a Tetris clone. (I'm not
being facetious here--I'm willing to bet that >99% of Norwegian cell phone
users will find Tetris more useful than a Japanese input method.) We may
start seeing multilingual-ready cell phone as a niche product sometime in the
future, when cell phones get big and fast enough to run the equivalent of
Windows 2000 comfortably--but probably not before. 

2) Good input methods are not trivial to write, and AFAIK, no decent input
method implementation exists in Java right now. Also, cell phone makers are
scrambling to cut down on development costs for their increasingly complex
systems-- witness their rush to embrace Linux. Put the two together, and guess
how likely it is for them to write a new input method from scratch in Java.

BTW, protocol-wise, Japan will become standards-compliant in a year or
so--DoCoMo and J-Phone will have finished extended W-CDMA coverage throughout
Japan, and KDDI already has CDMA 2000 1x running. The problem is with the rest
of the world, which has been dragging their collective feet in getting 3G
networks up and running. Privately, I'm starting to fear that 3G will never
get implemented worldwide, and Japan will be left with an incompatible
technology just like today.

-- 
Shimpei Yamashita                               http://www.shimpei.org/
You can't have everything. Where would you put it?    -- Steve Wright