Possible bug

[aldacron]

From Keyboard.java, starting at line 404 in enableBuffer():

readBuffer = nEnableBuffer();
if (readBuffer != null)
	readBuffer.order(ByteOrder.nativeOrder());
return readBuffer.capacity()/2;

The native implementation of nEnableBuffer calls env->NewDirectByteBuffer(), then returns the reference. If there really is a chance for a null reference to be returned, the the call to readBuffer.capacity() should also be inside the if block. I don't know if NewDirectByteBuffer would throw an OutOfMemoryError or not, but if it does then the null check is pointless anyway.

[Matzon]

yup, null check is superfluous since it will throw an exception if anything goes bad