[tlug] Re: UTF-8 Terminal Emulators?

[Mike Fabian <mfabian@example.com>]

"A.Sajjad Zaidi" <sajjad@example.com> writes:

> On Tue, Apr 16, 2002 at 01:28:24PM +0200, Mike Fabian wrote:
>> 
>> In the long run, UTF-8 is probably the only encoding which makes
>> sense. What else could you use to send multilingual email in a
>> standard way which everybody can read?
>
> That might take a while.

Yes, it will certainly a lot of time until UTF-8 becomes really
popular. But when it does, it will make many things easier.

>> With UTF-8, you can easily mix many languages, for example
>> 
>>     Japanese (こんにちは)
>>     German (Grüß Gott)
>>     Czech (Dobrý den)
>>     Russian (Здравствуйте!)
>>     Korean (???????????????,??????????????????)
>>     simplified Chinese (你好)
>>     traditional Chinese (早晨)
>>     ...
>
> I think everything except Korean showed up correctly,

Korean is destroyed in your reply although it was correct in my mail.

> although German and Czech had some large spaces in them.

This is probably a font problem on your side, it is still OK
in your reply. 

> So how does input work with UTF-8?

Most input servers work in UTF-8 as well. You can use for example
kinput2, Ami, xcin, and many others in UTF-8.

But with most applications you have to decide which one of these you
want to use at the time of the start of the application and you cannot
change that later.

I.e. you can do

    LC_CTYPE=ja_JP.UTF-8 XMODIFIERS=@example.com=kinput2 program

or

    LC_CTYPE=ko_KR.UTF-8 XMODIFIERS=@example.com=Ami program

or

    LC_CTYPE=zh_TW.UTF-8 XMODIFIERS=@example.com=xcin-zh_TW program

Usually you cannot change the input server in the running program.

There are exceptions though, for example the multilingual terminal
emulator 'mlterm' can switch between different XIM servers
(e.g. between the XIM servers mentioned above) on the fly at while
running.

IIIMF, which is supposed to replace XIM eventually, is designed to be
able to switch between different servers/languages at runtime, i.e.
when IIIMF becomes widespread, this should become easier.

And of course there is (X)Emacs. (X)Emacs has input methods for many
languages, can switch between them on the fly and works with UTF-8 if
Mule-UCS is used.

-- 
Mike Fabian   <mfabian@example.com>   http://www.suse.de/~mfabian
睡眠不足はいい仕事の敵だ。