How do you do it?

The standard ones ubuntu comes with just do not seem to work at all for me. I really need something that works the same as the Microsoft/Google IME on windows.

Anyone tell me?

Come tell us when you find out.

A few years back I tried to use linux, but had to give up when I realized it had no way (that I could figure out with my layman's knowledge) to input Korean.

http://code.google.com/p/ibus/
then install the anthy package, or search for the mozc package for the google input stuff

scim is also available, but its older and not as good, and doesnt work on X11 apps (urxvt, stuff running in wine, etc).

this also might be relevant.

What doesn't work for you about the Linux one (I assume you mean SCIM)? I have never used windows IME, what are you missing?

Japanese/Korean input works out of the box on recent Ubuntus but won't be available by default unless you log in with that language.

As for what Linux IMEs are missing, they are all pretty crap compared to the commercial offerings. Prediction is much worse and the dictionaries are lacking. If you do a lot of input in Japanese it's worth considering the commercial ATOK input system which runs on Linux, but it's proprietary and somewhat awkward to integrate into the system.

よし！ 出来た！

Thanks guys.

resolve Wrote:As for what Linux IMEs are missing, they are all pretty crap compared to the commercial offerings. Prediction is much worse and the dictionaries are lacking.

http://code.google.com/p/mozc/ is supposed to change that. I've been using it for a few months now and it's a whole different class than anthy.

It does look promising - I'll have to give it a whirl.

What, so I don't have to use anthy anymore?

...I love you guys. Time to vim some x configs.

I find anthy/ibus on linux more comfortable than the microsoft ime so far. Though that may change when my vocabulary improves.

じゃあ、また。

I use SKK, mostly out of habit these days. I do like the way it forces you to get the kanji-to-okurigana boundary right (you have to put in the capitals: KaKimasu => 書きます).

Linux support for input methods other than by setting the entire user locale/config to a particular language has historically not been particularly good, because most of the people who care about, write and document Japanese IMEs are in fact native Japanese speakers. So it can be made to work but can involve a lot of messing about under the hood if you're unlucky...

You need to get SCIM and ANTHY packages, by far the best IME I've ever used, I'm going to have to try this ibus though.

I'm running an old Fedora disto (F9) on this laptop. Only had to

# yum install scim anthy fonts-japanese

and maybe scim-lang-japanese and scim-anthy; I don't remember if anthy pulled those in by default or not

don't recall any configuration problems; didn't have to fiddle with locales; mine has always been set to the default (i.e., all LC_ envars undefined), and LANG=en_US.UTF-8

I've also installed some additional fonts as I've run across them, including the stroke order font and some true-type fonts

works great

I've installed tomoe and scim-tomoe for handwriting recognition, but I can't comment on the quality -- I've only played with it a little, don't really use it