* now the glm_mat4_inv_precise is deault, because I don't think all people are aware of this func * the old behavior (fast) replaced with new func: glm_mat4_inv_fast if fast version is desired then glm_mat4_inv_fast must be used.