Re: [tlug] Japanese Word processor for linux

[mune <fukudam@example.com>]

Shawn wrote:
> 
>> In my opinion it is a mistake to put Shift_JIS in java source.
> 
> Depends.  Your Keitai may not like Unicode.

Yeah, I agree with Shawn.

It is convenient using Unicode when making programs
in Java (especially in multi-language environment),
but it is not always the case that you can do this.
For instance, in my current project, database
is already configured with Shift_JIS, HTML source is
Shift_JIS, etc..  but somehow, the output HTML is
in EUC-JP. I'm not the one who set up this original
environment, and it's gonna be a great pain to set
all of these to Unicode now as data in database is
pretty huge already.

Also, I'm not so sure about major Java editors save
the source in Unicode..  I know Eclipse (it is a
pretty major one nowadays) on Linux saves Japanese
in EUC-JP by default, and probably Shift_JIS on
Windows.

Personally I think the choice of encoding for source
files doesn't really matter as long as everyone in
the project uses the same encoding.


 >> I also had the chance to work on a Java
 >> project with several people, one was saving
 >> stuff in utf8, the other was saving in sjis
 >> (obviously this was the windows guy), the
 >> third was saving it in euc-jp. Wasn't very fun.

Now this is ugly...


>> About the editor:
>> Don't listen to people who tell you to edit your source in openoffice.
> 
> Don't listen to people who tell you not to listen to people who suggest 
> a reasonable solution especially given that you probably have it and may 
> need to do something immediately.  Oh did you also catch the jedit 
> suggestion.

I wouldn't dare going into "don't listen to.."
argument, but editor and/or environment for developing
also shouldn't really matter as long as the code
created can be read by other ppl in the team without
any problem.

I've read a book on testing Java programs (I forgot
the title), and believe it or not, the author was
using Word. He gives an extensive explanations in
how to set up macros for convenient coding with
it..

At my current work place, most ppl use Linux (either
vi or emacs), one guy uses Mac OS X, and yet another
guy was using Windows (Eclipse). I don't think we
had any apparent problem in the past with this.

If there's a strict coding convention (e.g. usual
space/tab arguments..), maybe you can use
Jalopy (http://sourceforge.net/projects/jalopy/) or
Checkstyle (http://sourceforge.net/projects/checkstyle/).

I think both of these can be used as ant task, so
the source can be formatted before checking into CVS..


As for the original question posted by Sam (it's prob
already resolved by now), my Linux environment is set
up with Japanese during installation, so I'm not so
sure how to fix that...

It seems more of an environment issue of the OS than
setting javac options tho..  (maybe ppl at Ninjava
knows?)


Well, prob you (Sam) know about this already as I'm
sure you know alot better about Java than me, but
just in case..

If you want to compile the source files with Shift_JIS
encoding in Linux environment (default ja char is EUC-JP),
compiling the source files with "javac -encoding SJIS
[source]" should work fine.


-mune