> >I don't understand your point here. My point was that when checking for > >exactly zero, I would not expect my code to detect anything except zero > >as zero > > So, what about +0 and -0? +0 and -0 compare as equal in IEEE754.