> has to be done by something closer to the hubble telescope (pointed backwards). state of the art is about 2x diameter, with adaptive optics methodologies that avoid needing to polish a giga-mirror over several years