next up previous contents index
Next: Mule Input For Searches Up: Editors Previous: Editors



Editors in the Emacs family of editors are very popular in Japan. Nemacs stands for Nihongo Emacs and was first released in June 1987. Nemacs was a Japanese localized version of GNU Emacs and only supported Japanese and English. The last version of Nemacs was version 3.3.2 released in June of 1990. Future developments concentrated on Mule with the goal of support for multiple languages within the same editor, on the same screen, at the same time. Version 1.0 of Mule was released in August of 1993 and was based on GNU Emacs 18. The last version of Mule under active development was released in July of 1995, version 2.3, based on GNU Emacs 19.28. Future developments of Mule concentrated on GNU Emacs 20.x, which integrated Mule functionality into GNU Emacs.

There are two versions of Emacs in active use that support Japanese, GNU Emacs 20 and XEmacs 20. Both GNU Emacs and XEmacs must be compiled with Mule support. The default is to compile the programs without Mule support as there is a performance penalty for Mule support during runtime. Additional compilation options are needed for support of the various input methods available.

There are two binary distributions of mule, one for canna and one for wnn. In addition to mule, you will need to install the kanji conversion server. For example, if you install mule-wnn, then you must install jserver. The default install of Wnn comes with both jserver and the uum front end processor. You don't need to use uum from within mule-wnn.

This is an example of a linux-nihongo work environment centered around mule.

An integrated work environment centered around mule can still be utilized if you want to work with other programs. For example, if you like the xjdic dictionary interface, you can input Japanese using uum or kinput2. My recommendation here is to standardize on kinput2 for your front end processor. You can use kinput2 as the front end processor for mule if you run mule in a kterm. One drawback to this strategy is that you lose the ability to use the mouse in mule.

Mule, Wnn, and Tamago

Start the jserver if you have not already done so and call up mule under X with a reasonably sized font:

  cow$ mule -fn 10x20 &

The screen should be considerably wider than a similar window under English emacs. Remember that you are using double-byte characters now.

Switch to Japanese mode with


After switching to Japanese input mode by selecting

, Japanese can be inputted between two upright bars | | known as fences.

Type the roman charaters watashi. The characters should come up on the screen in hiragana. \includegraphics{editors/images/}

Press the space bar to convert it to Kanji. Keep pressing the space bar to bring up different Kanji choices until you find one you like. \includegraphics{editors/images/}

To input katakana while in the fences, first type the characters in romaji. The characters will appear as hiragana. \includegraphics{editors/images/}

Next, press M-k to convert the hiragana into katakana. \includegraphics{editors/images/}

Since many people have a dot between their first and last names, you might need to type the romaji characters ten. \includegraphics{editors/images/}Press the space bar to convert the text into the symbol. \includegraphics{editors/images/}Your name can be finished in either katakana, hiragana, or kanji. \includegraphics{editors/images/}

Another fence mode command is M-h to put into hiragana mode. C-c or C-g will cancel fence mode.

next up previous contents index
Next: Mule Input For Searches Up: Editors Previous: Editors
Craig Toshio Oda