The first is that if funding agencies, vendors and computer science researchers really want to help working scientists do more science, they should invest more in conventional small-scale computing.
See original here:
Mailund on the Internet » Blog Archive » How do scientists really …