If you want to try out new features, test if bugs are fixed, or just be on the bleeding edge, you can download our developers binaries. They can be quite unstable, depending on current development, so don't be supprised if your PDA crashes and needs to be reset when using one of them.

Latest Packages [2006-12-10 00:14] [2006-12-08 19:30] [2006-12-08 03:37] [2006-06-19 23:44]
More are in the padict nightly archive.
Stroke Database Updates [2007-09-08 20:02] [2006-11-08 03:58] [2006-11-07 04:21] [2006-07-17 11:08]
More are in the kanjidb nightly archive.
Unstable Dictionaries
Currently there are no updated dictionaries.


Warning! The latest PAdict code is not on the Sourceforge CVS server!

Instead, development is hosted on Sourceforge SVN. The following command will pull everything:

$> svn co

Information on Subversion and its clients is available at


There are many ways to contribute. You could help in making the core program better (faster, nicer, smaller...), help writing better documentation, or adding strokes for more kanji to the kanji database...

If you are interested in the actual, unstable development code of PAdict or the database generator code, both are in the CVS on SourceForge. The database generator is mostly quick&dirty C++, but seems to compile with Cygnus gcc under windows and GNU/gcc under Linux. I have only tested it on x86 machines, but the different data alignment on other systems may create corrupt output.

Database converters

If you want to convert your own edict formatted file for PAdict, you can use the padict dbcreator. It can be found in CVS, but we also have a windows binary:

Windows dbcreator binary

usage description

Don't expect too much, though, it's just a command line tool (yet)!

Developer resources

The Mailing List

The TODO list

I have made up some database descriptions. Those are only interesting for developers. I'm not quite sure that's it's possible to understand them, but give it a try ;)

developers_notes.txt - some global database formats and descriptions

structures_tree.png - how the index trees for kana and kanji are build and stored


Most of PAdict's dictionaries are based off the databases made by the EDICT project and Jim Breen. Thanks a lot for making this work available!

Ciaran Keating of Australia did much of the database search logic.

Once upon a time, JavaDict was created by Todd David Rudick. Robert Wells made jstroke based off this, helped by Owen Taylor. In turn, Ivan Kanis made PocketKanji based off these previous works, which PAdict adopted for handwriting recognition.

Stephan Matthiesen helped in revising the kanji recognition and making the database for kana recognition.

Kenji Baheux worked on Japanese PalmOS integration.

David Williams for par.

Some Japanese Palm Handhelds have been donated by Daniel Hauck to test Japanese PalmOS integration.

Alexander R. Pruss developed the palmresize code.

Eduardo Gonzalez has worked on keyboard support.

The website and documentation for PAdict 0.9.0 has been developed by Sudrien, with help from pretty much everybody contributing. Let's keep our Intarwebs clean, everybody.

Thomas Arendsen Hein, David Schneider, and Eric Kjeldergaard helped by bug-testing 0.9.0.

PAdict was created by Lars Grunewaldt. It is currently developed and maintained by Lars Grunewaldt and Benoit Cerrina.

This website and all contents, including the PAdict logo, are copyright © 2006 by The PAdict Project. All other trademarks are the property of their respective owners.