[Uim] FYI: anthy-5100

[yusuke at cherubim.icw.co.jp]

I have released anthy-5100 today. You can download it from
http://sourceforge.jp/projects/anthy/files/

Anthy is a Japanese conversion system and can be used
as one of uim's backends.

*Long standing bug in segment sequence learning was fixed.
*Prescision of conversion was improved by various fine tunings.
*Reduced memory consumption.
 (200KB to 130KB when there are no conversion context.)
*UTF-8 API was added.
 Currently, many of internal structures in Anthy depend
 on EUC-JP encoding. Probably, it will take 1 or 2 release
 cycles (x4 months) to support UCS natively.
*Preliminary syntactic structure analysis is implemented.
 (This actually improved stability and prescision of conversion)

P.S.
 If you have signed up orkut (www.orkut.com) community,
 please join Uim community too.
--
 CHAOS AND CHANCE!
  Yusuke TABATA (yusuke at cherubim.icw.co.jp)