So I need to be able to type Japanese. There seems to be a few different ways of accomplishing this, but I'm curious as to what people actually use.
Last time I talked to someone who had things configured, kinput2 and Canna was the way to go. Apparently this is old and busted now, and I should go with something else. But what? Just from random surfing, it seems that SCIM and anthy gets the most mention.
Ideally, I'd want something that is conceptually simple, ie does not consist of a million servers and programs talking to each other in mysterious ways. Easy to compile/install would be nice too, but that's not a requirement.
And if anyone here knows of a page that manages to explain how this all ties together behind the scenes, give me a pointer. Last time I tried (kinput2/Canna), I just ended up being confused and wanting a Mac.
Note: I don't run Linux, and not KDE or GNOME either. So most of the simple "just install $FOO package" solutions are out. (Sorry, guys.)
I guess im-ja is not an option, then?
That's the only one I ever managed to get working.
Looks like it would work, I do have GTK installed. But does "for GTK2" mean it has a GTK interface, or that it only works with GTK apps?
I think I'll have to read up a lot more before I can even begin to figure out how all this stuff ties together.
I use Anthy + UIM. It was fairly easy to set up: http://wiki.archlinux.org/index.php/Input_Japanese_using_UIM_%28English%29
The plugins it provides for gtk2 and qt work well and I've just checked that works also in generic X apps like xterm using uim-xim.
As for getting japanese input working in text console, I have no idea. I run linux in vmware and ssh to it using putty, so I can use the windows IME.
I just set up SCIM & Anthy some days ago and I guess it would have been fairly simple if I used either KDE or Gnome. It seems to work well with those desktop environments but it was a bitch to set up with fluxbox/enlightenment. The lack of any thorough documentation didn't make it easier either. It works fine now with most GTK/QT applications but it involved a lot of googling and experimenting.
So I guess what I want to say is: SCIM & Anthy is excellent once you get it to work the way you want but it is in serious need of better documentation. And some more general integration (not just KDE/Gnome) would be nice.
When I first got the Lemote boxes I have here, they came pre-loaded with Debian Etch localised in the Chinese language.
SCIM was the package installed for Chinese input. I know this as I remember clicking on the icon wondering what it was (the entire interface is Chinese, I don't speak a word of it) and working it all out.
こんにちは!
I went with uim and anthy, at least for now. (and mlterm)
Still have to figure out some things, like why the toolbar doesn't work, but it's good enough to be able to type phrases in the occasional email, which is about the level of sophistication I need right now.
Thanks for the pointers. And for anyone else reading, I found this to be at least somewhat helpful: http://en.wikibooks.org/wiki/Uim .
SCIM & Anthy for me. I'd still quite like to know if there's a way to cause the input method to interpret my western keyboard layout as though it were japunese; romaji is such a hassle when you've got used to a たていすかん layout otherwise.
OP here. There most likely is. At least with uim, there seems to be settings I can see (in uim-pref-gtk) to enter kana directly. ("Default kana input method" : "Romaji", "Kana", "AZIK") And there's this thing on the toolbar (uim-toolbar-gtk) that seems switch between them too.
Or that's what I guess it's for anyway. Cause it doesn't really work here -- I can either set mlterm to use "XIM" as an input method and have a functioning tooobar with all the mojibake I could ever want; or a working "uim" input method with a cold, dead toolbar.
Oh well. At least uim works with hotkeys.
I was surprised to learn that Anthy (in conjunction with SCIM) lets one type those "lower-case" symbols (i.e. ぁぇぃぅぉ) by prefixing the actual letter with a "l". Yet to figure out how to type syllables that don't occur in native japanese in katakana...
>>9
Found it -- SCIM/Anthy calls it "kana typing method". Now to learn the keyboard layout...
>>10
At least with uim+anthy, you get katakana when you hit F7. So I just type "te li - <F7> <enter>" for ティー (tea). (Yeah, silly example...)
Obviously it can be configured, but I think F6/F7/F8/F9 is at least kind of a standard.
Typing an unambigious syllable, like "fa", works like expected. (Interestingly, "va", etc, doesn't unless I manually extend the selection. Hm.)
If, like me, your needs are limited to editing documents and email Emacs is the easy way out. The built-in package GNUS can handle email. I use plan9port's htmlroff for documents.
I put this into my .emacs:
(set-language-environment "Japanese")
(prefer-coding-system 'utf-8-unix)
Then in any buffer, just press C-\ to switch to (I think) the SKK input method. M-x describe-input-method will give all the information you need. No OS customisation or further packages required.