I see, good point. I’m not convinced by this summary of the benchmark; N is too small, and they ignore the fact that views are much slower in Blaze than the oth

I see, good point. I’m not convinced by this summary of the benchmark; N is too small, and they ignore the fact that views are much slower in Blaze than the others. To be clear, I think it’s great to have multiple competitive C++ implementations out there.