Hacker News Reader

You don't need 100 iterations of binary search. Double-precision floating-point numbers only have 52 bits of precision.

Hey, thank you -- this is a good point. I'll update the code to 52 iterations in the upcoming iteration to the site.