> I'm not sure what you mean by the "basic comparison ops" but I'm > fine with the policy, handling each component values in the same > way with float. So point_eq_point() is right and other functions > should follow the same policy.
I mean <, >, <= and >= by basic comparison operators. Operators with those symbols are available for some geometric types, but they are comparing the sizes of the objects. Currently only the equality operators follow the same policy with point_eq_point (), others never return true when NaNs are involved. > Sorry for going back and force, but I don't see a difference > between it and the original behavior from the point of view of > reasonability. Isn't is enough to let each component comparison > follow float by redefining FPxx() functions? My previous patch was doing exactly that. I though that is not what we want to do. Do we want box_overlap() to return true when NaNs are involved?