Version 1.66 has been released.

The mailing list has been decommissioned. Apologies to anyone who was depending on it.


This is a greatly enhanced fork of the defunct JWPce Japanese text editor (described below). Many bugs have been fixed and plenty of new features have been added.

I have named it JWPxp due to the removal of support for WinCE and DOS-based Windows (e.g. Windows 9x) as well as to differentiate it from Glenn Rosenthal's original product. The ur-program was called JWP. Later, the "ce" was added to the name by Rosenthal because he added support for WinCE (more or less completely rewriting the program in the process). Now I've changed it to "xp" because Windows XP is the lowest officially supported OS, although it probably works fine on Win2k and possibly even NT 4.x. Maybe by 2025 we'll see a version called JWPw7 or JWP7.

Introduction to the Progenitor

JWPce (the precursor to JWPxp) is a professional, open-source, Japanese word processor that originated in the late 1990s. In addition to word processing / text editing capabilities, it also provides numerous kanji lookup tools and integrated dictionaries. It runs on most versions of Windows, including WinCE, from which its name derives. The last update was released in 2005, and the official webpage went offline in 2015.

It is an old-school Win32 program written in C++ for the MFC framework. It does not have a "ribbon" interface or eye candy in its GUI, neither does it utilize .NET. Although it has always been free, its feature set is commensurate with commercial-grade products of the time.

At one time it was successful enough to merit a Wikipedia article.

JWPce has been largely superseded by JWPxp, although it might still be useful for users of older operating systems. You can download it below.

Prominent Enhancements

  • Unicode support: files with non-ANSI characters in the name or path can now be operated upon
  • Can now receive Unicode and IME input (requires a system configured for IME - toggle with Ctrl-M)
  • Improved speed when working with huge files such that you can even manipulate EDICT directly
  • Will now display the largest EDICT entries (suru, tsukeru, etc.) without truncating them
  • Better dictionary searching, particularly for advanced searches
  • Single-kana searches are now possible (useful for looking up particles)
  • Greatly expanded sorting abilities for dictionary searches
  • Numerous usability enhancements and additional keyboard shortcuts
  • Brace matching (for source code)
  • The binary configuration file has been abolished in favor of text-based
  • Lots of other improvements as noted in the change log

  • Dozens of bugs have also been fixed. It should no longer crash on any dictionary search, for example.


    Q: How often are updates released?

    A: Assuming the number and severity of reported defects remains low, it depends largely on the vagaries of my interest in developing the program. Typically there will be many months or even years between releases. Bug fixes and minor changes tend to accumulate until there are enough to justify a new release.

    Q: When will all the bugs be fixed?

    A: I naively figured that a mature program such as this would not have too many bugs and that a state of near-perfection could eventually be reached. Despite all the progress I've made, that theory has not borne out. If you want a particular bug to be fixed quickly, you should report it, thus guaranteeing my awareness while simultaneously generating urgency to act.

    Q: May I submit a feature request?

    A: Sure. It may not ever get implemented, but it will at least be considered. Simpler requests have a higher probability of success.

    Q: What about source code changes?

    A: Please send entire source files, not diffs or patches. Make sure your editor is not inserting tab characters. Indentation is 2 spaces. (I don't like it either, but that's the way it was originally written!)

    Q: Any chance of resuming support for such-and-such configuration?

    A: I would be delighted to continue supporting Windows 95, obsolete PDAs, and other platforms; I just don't want to do the maintenance (compilation and testing, mostly). I have not removed the conditional compilation statements and I've even added a few for Unicode-specific sections of code I introduced (although ANSI support might not even be necessary anymore as I vaguely recall some sort of Unicode layer for Win9x and possibly WinCE). Please coordinate with me if you are interested in accepting responsibility for one of the formerly supported environments.

    Q: What's a good way to learn to use this program?

    A: Read the help file. Though outdated and slightly inaccurate at times, it's still highly relevant and very thorough.

    Q: Why would I want to use this program?

    A: Some of the features I consider to be the most useful are listed below. I used JWPce for most of my translation work, and at least one commercial company has used JWPxp for translation work.

  • Integrated dictionaries, including user-defined entries
  • Fairly powerful dictionary search features, including automatic deconjugation
  • Numerous kanji lookup methods (and one built-in input method)
  • Ability to read/write several common Japanese encoding standards (Shift-JIS, EUC, UTF-8, etc.)
  • Portable and completely self-contained — no need to install anything, including the program itself (*)

    * Optional IME support is provided by the OS and therefore may need to be installed separately

    It also has other features I have limited use for, such as printing, page layout, colorized kanji (a learning aid), statistics, and the ability to export text in graphical form.

    Little-Known Licensing Facts

    This software is licensed under GPL Version 2, but you are not required to accept the license in order to merely download or execute said software. However, nothing else grants you permission to modify or distribute the software or its derivative works, and such actions are prohibited by law if you do not accept the license.

    JWPxp Downloads

    JWPxp complete package as a single 7-zip archive

    JWPxp source code

    Don't forget to update your dictionary files periodically! I only include the original 2005 dictionary (from the final JWPce release) in the distribution archive since it's good enough for most purposes and saves space. Besides that, EDICT is updated constantly so it is impractical to keep current. Be aware that ENAMDICT is not included so if you want to look up names, get it below and unzip the files into the same directory as the JWPxp executable.

    EDICT (latest version, with index)
    ENAMDICT (latest version, with index)

    The main EDICT page has additional documentation and related files. You have to do quite a bit of digging to find the indexed ENAMDICT, and it can be confusing to decide which files to download from that page, so you're generally better off using the links above if you only want to update your dictionaries for use with JWPxp/JWPce.

    Old Downloads

    Here is an archive of the original (unsupported) JWPce for historical or comparison purposes. You probably don't need this unless you are running a very old Windows OS. The complete source code for this version is included with the JWPxp source code above.

    JWPce complete package

    Technical Support

    Limited technical support is available via my gmail address. Please allow up to 2 weeks for delivery. The user name to which any inquiries should be sent is 'mijeto'.

    Final Words

    I've held a longstanding skepticism towards the GPL. This is the first time I have released any of my own source code under it. Please don't make this a negative experience for me.

    Thank you, Glenn Rosenthal, for your hard work on this excellent program.