And it is hard to apportion blame/credit here. In my experience, the wins that I like the most aren't the most effective ones. I suspect that almost all of this win was really just gotten by not passing around as much of the graph.
On Thu, Jul 1, 2010 at 11:43 AM, Ken Krugler <[email protected]>wrote: > All in all, seems like he wound up with a 3x speed win (31% of original > time), IIRC. >
