what is TODO. Please stick to this if you want to please me with 
your developments. If you want to please yourself, do whatever
you want. Any enhancement to PAdict is a good enhancement, whether
it was my own idea, or not :)

* and there will always be the code *

The TODOs I consider MOST important are [11],[13],[7],([4],[9])

PAdict/Palm-stuff
-----------------
Stuff that has to be done for PAdict itself

* search algorithms 
  The index of PAdict for kana and kanji opens many possibilities for
  searching. It's possible to search for nearly anything, even fuzzy
  (ignoring spell errors) or starting mid-word. Right now, these options
  are not even far from being implemented.
1]- be creative here :)
1a- of course, search speed could be improved. most things are not handled
    very sophisticated in any way (memory usage, CPU usage, ...)

* database handling
  Right now, only one database can be used. In the future, it should 
  be possible to install more than one database. Also, databases should
  be able to grow past 65535 words.
2]- enhance the database handler to search for all databases installed
3]- enhance the database handler to take care what results are found in
    what database (i.e. the references in hotlist must contain the dict 
    they came from)
4]- enhance the database structure so that large indicies (32 bit instead of
    16 bit) are used for result handling and that the index mechanism works
    for search trees that are broader than 64k nodes (needed i.e. for a full 
    jmdict)
    This should be ENHANCEMENTS; it must be possible to have both database
    types installed at once (because the small database indicies should not
    be blown up)

* trainig module
5]Well, be creative here, too. Result storage, intelligent (card based) 
  questioning, ...

* pocketkanji
6]Stroke recognition could be possibly improved. Kana recognition MUST 
  be improved

* kanji definitions
7]Add SODs if possible to gain such data from somewhere (Jack Halpern or
  mjordan (at) code-fu.de)

PC/Tool-stuff
-------------
Stuff that needs to be worked out but does not concern the
Palm app itself

* database generation tools
  There are some tools that convert the font and dictionary files to
  a format readable by PalmOS (.pdb format). Those tools mostly do not reside
  in the PAdict source, but are kept in their own realm. Check out
  padict-dbcreator 
  from the CVS. The tools are all VERY quick&dirty, even if I tried to use C++
  wherever possible. It's not even documented very well. The font converters 
  are all right as they are right now, but the database converters should be
  enhanced.
8]- XML support for JMdict instead of EDICT
9]- support for large databases ( > 65535 words) [needs PAdict enhancement]
10- better code structure ("plugins") for more import modules to cope
    edict, jmdict, fjdict (edict format ja->fr), wadokuJT, ....
11- word selection method based on word usage  in japanese (most used words 
    to create smaller dicts)
    It should be possible to gain this data by digging japanese web pages or
    novels for words, but it might be a hard work
12- cross-platform frontend so that all people can use it
13- create more MORE MOOOORE dictionaries for small PDAs, big PDAs, red PDAs,
    blue PDAs... hrhr ;)





***

old TODO file contents, should be removed someday :)

just some suggestions what should be done...

redesign:
- program state should be stored in preferences
- screens should use the same (or mostly the same) menu

major features: