[Opensg-users] Precision of Quat.setValue with doubles

[Allen B. <abi...@gm...>]

I just ran into an interesting case where I have two double Vectors and I
try to pass them into the Quaternion setValue method that takes a rotateFrom
and rotateTo.

You can see the code here:

http://www.opensg.org/doxygen/classOSG_1_1QuaternionBase.html#a64116d4041c7f2c2e7500f1ac37d99a5

The issue is that the tests against the cost variable for degeneracies isn't
precise enough for doubles and considers the vectors to be the same even
when they are not.

I suggest adding 4 or 5 more 9's on the test here to make sure it can work
fine with double values.

-Allen