[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Debian m17n method under X with UTF-8


On Sat, Oct 16, 2004 at 11:50:18AM +0800, Arne Goetje wrote:
> On Saturday 16 October 2004 07:34, Osamu Aoki wrote:
> > How are you making Debian into UTF-8 and multilingual/multi-locale
> > system while switching between normal gnome and Xfce? So far, I did
> > not find this trivial.
> Well, in my case, I found SCIM to be the best solution. 

SCIM does not address needs to access traditional consoles. So this
multilingual environment problem is not just easy IM.

I agree SCIM is a good option and that is why I sponsored its simple bug
fix upload recently.

This package "m17n-env" is about offering access to the traditional
consoles too.  (Also giving option to choose traditional IM too.)  You
know, Debian is choice.  Also writing documentation for UTF-8 conversion
is not as clean as telling to install helper package to do it instantly
:-)  If you are using .xsession to do your customization, GDM desktop
environment choice may result skipping your SCIM setting under Xfce or 
KDE. So there are many aspect to it.  If you get into emacs settings
like language-env package, the issues will be much bigger but I did not
go that far.  Again, it is in

  deb http://people.debian.org/~osamu/package/ ./

Also post-1.0 version of SCIM needs to be packaged with scim-uim and
others. It is on-going task. If you can help, please drop me a personal
mail.   (I may drop experimental compile of SCIM there soon, too.)

> I don't need to care about locales any more, because SCIM doesn't care
> about locales, as long as you use any UTF-8 locale as default. 

Maybe you do not need it.  But there are many data in non-UTF-8
encodings and editor access to them is essential for the smooth

> If you want to try SCIM, compile it from source 
> (www.freedesktop.org/Software/scim/), the debian packages don't work 
> for me.

... well you need post-1.0 program sets for some features.  Some API 
has changed.

> The SCIM approach is in my opinion the only useful one, since the locale 
> should only reflect which language you prefer and in which country you 
> are. It should have nothing to do with input methods, as long as your 
> system is UTF-8.

That is sometimes not an easy option for editing traditionally encoded 


Reply to: