maybe i've spent too much time trying to make computers operate efficiently, but it strikes me that if a process a) takes more time than should be necessary and b) produces sub-optimal results, we should maybe pursue other processes.
and stop paying these idiots 7+ figures.