A cool recent blog post about not using epsilons in floating point math https://lisyarus.github.io/blog/posts/its-ok-to-compare-floating-points-for-equality.html

Guess this bug kind of proves the point