Bukkit

• fixed a bug where the hfillmap[i][xyz] could be indexed with a negative Y component (below zero) which would cause the thread to throw an exception and stop working on the worlds assigned to it and lock up the main bukkit thread

The chunks are regenerated almost perfectly except when water is the top layer. If water is the top layer some (a few) water blocks are missing. Also, moderate debugging output is enabled. This is nice for tracking where it is working at and being able to teleport there and view the process happening.

This appeared to be a problem, but has not shown up as much as I expected or none.