Hmm. The design is harder than the implementation here... how should we calculate the scaling factor for any commit? If we start off with only small commits, should the scaling highlight minor variances in size? And then if we discover a whopper, should the scale automatically recalibrate to keep the maximum node size capped?
Hmm. The design is harder than the implementation here... how should we calculate the scaling factor for any commit? If we start off with only small commits, should the scaling highlight minor variances in size? And then if we discover a whopper, should the scale automatically recalibrate to keep the maximum node size capped?