Wow you're quick. Thanks for the benchmark.
I guess 'num-sort' is not that much faster than 'num-ST' because in both cases there's still one (but only one!) round of stringification being performed per sub call (explicit pre-stringification in num-ST and implicit, sort of in-place stringification in num-sort).
In reply to Re^2: length() and the Schwartzian transform
by hazylife
in thread length() and the Schwartzian transform
by hazylife
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |