Why the hell does BoomerAMG have such shitty scaling on the Laplace equation: http://guest.ams.sunysb.edu/~zgao/work/airfoil/scaling.htmlCan we rerun this on something at ANL since I think this cannot be true.