Javascript floating point number usable ranges

Question

I finally got my infinitely zooming grid to work and tried out how much I can zoom. Everything worked fine until around 15,000,000,000,000x magnification where it got jerky and didn't move right anymore so I figured I had hit the limit of precision of whatever variable type its using. Then I tried zooming out and expected a similar limit but 5 minutes later, at magnification of 0.00000-insert 100 zeros here-0014829095701452545 it was still doing great. If 1 was 1m then that in relation would be Planck length inside a Planck length inside a Planck length. Crazy!

So not really a problem question but a curiosity one - What kind of types does javascript use for floats? and why is it capable of using much longer tiny numbers than big numbers?

Did you try several interpreters and see the same behaviour? — Chris Wesseling
– Chris Wesseling, Commented Jun 22, 2014 at 22:01

Niet the Dark Absol · Accepted Answer · 2014-06-22 22:03:34Z

2

The problem is where the precision loss is.

When you have a huge number, you start losing precision around the units, which can cause huge problems.

With a really tiny number, your precision loss is at very low decimal places, where you'd have precision loss due to not having enough digits to store the number anyway. So there's no real problem here.

Does that make sense? Basically it's the difference between:

1234567890123457000 -- big precision loss (321)
0.00000001234567890123457 -- tiny precision loss (0.000000...0001)

answered Jun 22, 2014 at 22:03

Niet the Dark Absol

326k86 gold badges480 silver badges604 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Javascript floating point number usable ranges

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related