In CompSci it's pretty rare for there not to be open access versions of a major paper. Off the top of my head there have been a few in Nature but outside that almost everything ends up as open access.