* Performance improvement for the deck editor when scrolling cards (transform 2*N*log(N)<Cmp1> + N*<read> into log(N)<Cmp1> + N*<read>). I expect this improvement to be much more significant, performance-wise, than last night's one.