From xemacs-m  Thu Mar 13 01:48:05 1997
Received: from jsk.t.u-tokyo.ac.jp (monoceros.jsk.t.u-tokyo.ac.jp [133.11.96.252])
	by xemacs.org (8.8.5/8.8.5) with SMTP id BAA19713
	for <xemacs-beta@xemacs.org>; Thu, 13 Mar 1997 01:48:03 -0600 (CST)
Received: from jsk (localhost [127.0.0.1]) by jsk.t.u-tokyo.ac.jp (8.6.12/3.4Wbeta6-96040821) with ESMTP id QAA18444 for <xemacs-beta@xemacs.org>; Thu, 13 Mar 1997 16:47:58 +0900
Message-Id: <199703130747.QAA18444@jsk.t.u-tokyo.ac.jp>
To: xemacs-beta@xemacs.org
Subject: Re: Japanese input (was Re: gnuattach) 
In-reply-to: Your message of "Thu, 13 Mar 1997 16:13:43 JST."
             <199703130713.QAA00713@pentagana.sonic.jp> 
Date: Thu, 13 Mar 1997 16:47:57 +0900
From: Hajime Saitou <hajime@jsk.t.u-tokyo.ac.jp>


> >>>>> "Steve" == Steve Dunham <dunham@dunham.tcimet.net> writes:
> 
> Steve> I also have a completely tangental question: I'm planning on
> Steve> learning japanese this summer and wondered what input method is
> Steve> recommended with xemacs-20: wnn or canna?
> 
> Of course I'm going to tell you wnn ;-) It still has some bugs, but
> those will be fixed Real Soon Now (as soon as the current crunch here
> at the office subsides). Actually, either would probably work well,
> with the folloing caveat emptor. canna is completly contained within

Wait, Canna needs a server too. You need to boot up Cannaserver. But
yes the installation is easier. You can also get prebuilt versions
of both programs from the JE distribution if you use Linux.

> xemacs, while wnn relies on jserver, an external process. And without
> some patches, jserver doesn't quite work out of the box. I'll be
> postingg some patches on Saturday for wnn, Xlib and (hopefully) wnn
> and after that you can make a better choice, perhaps.
> 
> Good luck on your studies!
> 
> --Jareth

The program that converts Kana to Kanji is called the "front end
processor."

There are other minor differences when you use them. With Wnn you can
temporarily stop the front end processing to input english words. You
can't do this in Canna. Though this may be due to my setup.  On the
other hand, Canna supports automated conversion. Usually, when you
input Japanese you have to press "Space" for the actual conversion
from Kana to Kanji to take place. With Canna you can just keep on
typing and the program will convert without your input when it has
enough information for a relatively accurate conversion. This may not
be that helpful as you usually have to redo a lot of the
conversions. It is getting better though. Canna supports a reconvert
feature too. This infomation may not be that useful now, but you'll
understand once you start writing a lot.

There are also ways to use connect to commercial front end processors
for DOS running on dosemu if you use Linux or FreeBSD. I wonder how
difficult it is to add it to XEmacs.

2.01 betas up to b6? Seems I've got a lot of catching up to do...
Keep up the good work!

Hajime Saito

