Discussion and Conclusion

After these presentations, there was a brief period of open discussion where the participants tried to sort out their thoughts and decide whether a clear winning theme had emerged. There was substantial sentiment that the concept of ``No-Futz Computing'' was a winner. The discussion then turned to how we could address this challenge. One school of thought, represented by Armando Fox and Margo Seltzer, felt that there was insufficient market penalty for low FutzMark ratings and that the immaturity of the field was the source of this problem. A second school of thought, represented by Jeff Mogul, felt that system management was primarily about policy, not mechanism. Mogul remarked that it is all of the things you have to do when the computer can't adapt itself to you. As such, it is not easily automatable and is really a social, not technical, problem. This led to a heated and confusing debate, involving Margo Seltzer, Mary Baker, Armando Fox, Stefan Savage and a number of others, on how complexity could be reduced in systems. No real answer emerged, though there were many clever repartees.

In the waning moments of the workshop, the participants were asked for a show of hands on whether ``No-Futz Computing'' emerged as the dominant problem from the breakout session. A substantial majority of hands were raised in response, and so this was declared to be the clear winner. In response to a request from Randi Thomas, the participants were also asked for a show of hands on whether ``Post-PC Computing'' emerged as an important problem. Many hands were raised, but distinctly fewer than for ``No-Futz Computing''.

In closing the workshop, Satya requested a final round of applause for the excellent leadership of the General Chair, Peter Druschel. He also noted that the true measure of this workshop's success would become apparent only a few SOSP's or OSDI's hence, when the ideas that were debated here emerge at the core of high-quality research publications on innovative systems.