Hello.. my name is Rob and I wrote the Nihongo Flash.
Yes.. there are 10,000 vocabulary expected for JLPT 1.. however.. the official lists show around 8000. The other 2000 are up to the test maker to pick.
I parsed my lists from various lists on the net and cleaned them up and removed double entries (that I could find). Levels 3 and 4 were from jlptstudy.com and levels 1 and 2 were from
http://www.thbz.org/kanjimots/jlpt.php3. Levels 3 and 4 were gone over manually and seem to be completely correct. Levels 2 and 1 were a bit trickier. Level 2 seemed to have about 1200 too many vocab. I searched for entries that already existed in 3 and 4 and this removed 1000. I trimmed the definitions to be a max of three.. (instead of 7 or more like the original data). I them took EDICT sub (22,000 of the most common words) and removed any that I already had in other lists.. this left me with an extra list with about 15,000 words.
Anyway.. I believe I have made a set of lists that is consistent and clean. Hopefully.. somebody somewhere someday will go over levels 2 and 1, make sure there aren't mistakes or double entries and post them back to the net. It is kind of hard to find good lists. And many sites don't give you machine readable sets..
Also.. there is a Java version underway. I have already put the base project onto sourceforge.. however it is just the UI.. no flashcard logic has been added. I would think about two weeks.. maybe one.
If anyone sees any bugs or errors in the lists.. please let me know.. the contact info is at the top of the dataset files.