In order of priority:
1. Fix any bugs.
2. Add option to produce all permutations of anagrams.
3. If anyone thinks it is a good idea, I can add the option to remove
   all single character words except "a" and "I", I have noticed
   that some unix dictionaries contain an entry for every letter
   in the alphabet, this produces many useless anagrams.  Perhaps "U"
   should also be left in, does anyone have any further suggestions?
4. Find some way of finding semantically/syntactically meaningful anagrams,
   probably using one of the publicly available tagged dictionaries.
5. Add a GUI.
6. Repeat 1.

If anyone would like to help with any of the above, please mail me at:
richard@deep-thought.org