orientate3 vs. orientate4

[mortennobel]

There seems to be an inconsistency between orientate3 vs. orientate4 (eulers_angles.inl). Note that the parameters of yawPitchRoll is different. I believe the orientate4 is the correct order.

template <typename T, precision P>
GLM_FUNC_QUALIFIER detail::tmat3x3<T, P> orientate3
(
    detail::tvec3<T, P> const & angles
)
{
    return detail::tmat3x3<T, P>(yawPitchRoll(angles.x, angles.y, angles.z));
}

template <typename T, precision P>
GLM_FUNC_QUALIFIER detail::tmat4x4<T, P> orientate4
(
    detail::tvec3<T, P> const & angles
)
{
    return yawPitchRoll(angles.z, angles.x, angles.y);
}

[mortennobel]

I would guess that both approaches are wrong. Shouldn't the correct solution be yawPitchRoll(angles.y, angles.x, angles.z);?

[Groovounet]

You are right orientate3 is broken. This is now fixed in GLM 0.9.5 branch.

Maybe logically the function should be "yawPitchRoll(angles.y, angles.x, angles.z);" but the implementation work in reverse order...

Thanks for contributing,
Christophe

[DashW]

Still confused as to why orientate4 is calling yawPitchRoll with the parameters ZXY... the function eulerAnglesYXZ has exactly the same implementation as yawPitchRoll, so does that expect the same ordering as well? Something's not right here...