3 ms·
This approach, at least in languages like C and Java, would effectively smush the effective MAX_INT down to MAX_INT/1e15 or whatever. No bueno.
by sk5t 8y ago
This approach, at least in languages like C and Java, would effectively smush the effective MAX_INT down to MAX_INT/1e15 or whatever. No bueno.
- strainer 8y agoIt doesn't because basically - the point floats. The max_int doesn't mean the largest whole number that can be represented, it effectively means how much resolution you have on the number line in every one of a very wide range of scales. These languages all give exactly the same standard IEEE754 float math results for basic operations.