glm_vec_scale multiplies vector (all items) by a scalar
result = vector * factor, new scale = ol...

* now we can include library headers like this: #include <cglm/cglm.h>

I think _copy is better and COMMON name for duplicating/copying
vector, matrices

* move SIMD instructions (SSE2, AVX) to arch folder

* first normalize rotation vectors because we only want to rotate here
not scaling

we can avoid function calls with this func

this is may useful for rotate a vec3 with affine transform

because v and dest may be same vector

* when window resized we only change aspect ratio, so projection still
keeps same