Quaternions are obsolete math, due to modernized vector math.

Not trying to flame, but ive been learning a bit of math lately and have been checking on a few things around the forums too.

This reply might solve more than one issue on multiple threads...so ill just say it here in the sticky one...

I think i found a slight issue in the math and might know how to fix some issues.

Make it known that this same problem is on multiple threads in multiple libraries.

Sidenote: OTHER libs, Magnitude is not correct, this quat lib one is fine.

Here, In the normalize function....the output is always less than one...almost half even. eg ||V|| = <1

Normalizing, mathematically speaking(for everything else to function properly), the math requires a normalized vector to have the magnitude of 1. eg. ||V|| = 1

unit vector = vector * ( 1 / mag )

as opposed to

unit vector = vector * mag

as it were.

That said, you can make a scalar function too....instead of 1 / mag make it Value# / mag, after you normalize ||V|| = 1

This will set the absolute size of the vector, no need to scale it up of divide it down.

However, those 2 scaling functions(scalar divide and multiply) are also possible, as well.....though this math is obsolete.

At least update the normalize function, so we got it right.

Love ya Ian, and thanks for such an awesome lib!

PS. if anyone has vector math libs( a couple of you), check your magnitudes, and normalizing again...itll fix crashes, divisions of zero, collapsing numbers, the need for a "solution", and no need for tolerances of numbers, nor rounding the precision, etc etc etc!