I agree about CS reviewing for conferences (esp. in systems) - pretty reasonable for the most part. Also, if things just don't work, they don't work - which is

I agree about CS reviewing for conferences (esp. in systems) - pretty reasonable for the most part. Also, if things just don't work, they don't work - which is a bit different than studies in social science.