Guerric writes: > I don't know if there is algorithms really dnet-friendly. The algorithms have reasonable parallel scaling from what I've seen. Biggest problem is ALL the results contribute to a success, making the prize funding much different than "winner takes all". John