3 ms·
I get you point. I'd probably go float/double precision for that task.
by quickthrower2 8y ago
I get you point. I'd probably go float/double precision for that task.
- pvg 8y agoThen you can lose precision that you might actually need - a bug that's even more annoying to track down than straight overflow.
- quickthrower2 8y agoYes indeed! But what is the alternative, especially if you are using approximations to transcendental constants?