11:49:09 For the contest, I'm glad to read in an original draft of the rules : "Submissions must score at least 20% better than the reference." Though I'm curious if creating a more nuanced scale would be beneficial. E.g., if a submission is 100% better, then the award is 100% of the available reward, and if the submission is 50% better, then the award is 50% of the available. Obviously n ot those exact numbers.... we'd have to guess what a reasonable maximum is. But this might encourage people to really squeeze the most out of the code, as opposed to getting 20% and going "welp, good enough for the prize". 11:51:14 also creates the possibility to keep the contest rolling, as the "reference" would change as submissions are accepted and adopted (or something) 11:58:10 probably also need to specify a reference machine 17:00:57 what if only suckers feel encouraged by such "competition"? 17:05:34 Then the competition will suck