There are several ways to interface with the dictionary files. Some of the more common interfaces are:
I use trans, trans.el, and xjdic on a daily basis. I have
mule start up on my X11 desktop. I then use mh to read my
e-mail within mule. If I receive Japanese e-mail and I
want to translate a few kanji into English, I type C-c C-t
to call up trans.el which calls up trans and loads the
EDICT dictionary. This takes up about 20 megs of RAM on my
system. Once the dictionary is loaded translations are rapid.
I then read fj.os.linux with gnus. I follow the same strategy
with kanji translation. If I want to respond to a message, I
load jserver if I'm using Wnn and activate Japanese input
C-\. If I need to check my spelling, I often use
xjdic to input the English word and get the Japanese translation.
Xjdic gives more choices on the translation.
I then read some of the top stories on
Nihon Keizai Shimbun
I find that using trans makes reading the newspaper a pleasure
instead of a tedious exercise.
If I need a break from technology and business, I read some
of the Japanese online novels that can be found at
Japan Open Yellow Pages Novel section
This type of casual relaxation
would not be possible if I didn't use trans and EDICT under Linux.
Xjdic has more features than trans and I enjoy using it as a standalone dictionary. I find it useful for translating both hiragana into English and English into Japanese. I also use it when I copy and paste Japanese from netscape.