Hector Martin

This is about string representation; I've explained this in other threads. JS has one number format, but it *pretends* to make the distinction when converting to string, by notating floats with exact integer values differently (without a decimal point, unlike other languages).

Which is all fine and everything, until it starts spitting out decimal representations of integer values that aren't the actual integer value represented by a given float.

My point is that floats *are* Proper Integers: up to 53 bits they can represent every possible integer, and beyond that they can represent certain integers (but those integers, they represent exactly). JS is printing a *different* integer than what the float represents.

Sure, it's repeatable, and it round-trips, but it's just /wrong/. If the float has a given value, and you print a *different* value, as an integer (not a decimal number that is obviously rounded), that's just messed up. Sure, you can *define* it that way, but still.