Change Details

There was added a camera test in rP22547. It was a strict comparison of two `CVector3D` arrays. Which means each float component was going to compare directly as well. There is an example of the failed test in comments of rP22547. Also you might notice that these float numbers differ only in floating point steps. So it might make sense to compare floats with EPS, since we can have some precision errors during few multiplications.

There was added a camera test in rP22547. It was a strict comparison of two `CVector3D` arrays. Which means each float component was going to compare directly as well. There is an example of the failed test in comments of rP22547. Also you might notice that these float numbers differ only in floating point steps. So it might make sense to compare floats with EPS, since we can have some precision errors during few multiplications. `EPS = 1e-4` is used because it's minimal possible EPS for such float values. Maybe it make sense to increase it to `1e-3` since it's the pretty good precision still.

There was added a camera test in rP22547. It was a strict comparison of two `CVector3D` arrays. Which means each float component was going to compare directly as well. There is an example of the failed test in comments of rP22547. Also you might notice that these float numbers differ only in floating point steps. So it might make sense to compare floats with EPS, since we can have some precision errors during few multiplications. `EPS = 1e-4` is used because it's minimal possible EPS for such float values. Maybe it make sense to increase it to `1e-3` since it's the pretty good precision still.