```(0.0).equals(-0.0) // false (0.0) == (-0.0) ...

# announcements

[Yossi Saiada]

(0.0).equals(-0.0)  // false
(0.0) == (-0.0)     // true

Does anyone can explain it?

[Pavlo Liapota]

Similar case was discussed yesterday.
Seems like runtime vs compile time calculation.
Report a bug.

[ghedeon]

https://kotlinlang.org/docs/reference/basic-types.html#floating-point-numbers-comparison

[ghedeon]

it's not a bug

[commanderpepper]

"-0.0 is considered less than 0.0" from ghedeon's link

[Pavlo Liapota]

@ghedeon thanks for the link!

[ilya.gorbunov]

Yeah, @ghedeon is right, it behaves as specified.
The bug that was discussed earlier is about double comparisons in constant expressions: https://youtrack.jetbrains.com/issue/KT-30282

[Yossi Saiada]

@commanderpepper if so, why does (0.0) == (-0.0) return true?

[ilya.gorbunov]

The operator == uses IEEE-754 order, which considers them equal.
The function equals uses total order, so these values are distinct.

[Shawn]

So, does == not actually always translate to a .equals() call?

[ilya.gorbunov]

Yes, it's the special case for floating point numbers: https://kotlinlang.org/docs/reference/equality.html#floating-point-numbers-equality

[ghedeon]

Right, not for these types. In this case equals behaves differently to support generic use cases and provide total ordering. So, it's good for sorting and stuff.