Earlier quoted context omitted.
Throwing 2d10 of different colors is equivalent of trowing 1d100. It's nice they have different colors to avoid discussions, but you can throw them in two different bins or one at a time or something. Remember to sum them as (x-1) * 10 + (y-1) + 1, that is a clear indication of why zero-based indexing is better. (Does someone sell "decade" dice, which faces say: 10, 20, 300, ..., 90 and 100?)
> (x-1) * 10 + (y-1) + 1 Is that not equivalent to: > (x-1) * 10 + y or: > x * 10 + y - 10
And some changes may have to been backported, and it has a lot of tricks with index of arrays of different dimensions, so I'm wrapping the formulas with +1 and -1 and hopping the best.
IIRC the python compiler does not optimize them (perhaps with numba?), but later steps in other programs are slow, so N [1] If the file says "1 2 7.0 \r 1 2 8.0 \r" should I keep the sum (15.0), the first (I never seen that) or the last? (Raising an error, nah.)