@johnolilly @donlbe ars is ok but basing all your "performance" evaluations on a run of Geekbench? Lame.