Charlie Catlett wrote recently about the GENI program. A quiz: is this ambitious proposal from the networking research community intended to (a) redesign the Internet, (b) get a lot of new money for computer science research, or (c) build testbeds to support innovative networking research?
The answer is "all of the above." Thus, I find GENI interesting for several reasons. First, there's the research agenda, which is to redesign the core Internet protocols to incorporate security, quality of service, and other good things left out by the original designers.
Second, it's interesting to see the computer science community making the case for a large infrastructure project. Historically, other sciences have done "big science", while computer science has emphasized small-scale, single-investigator research. GENI's supporters argue that a "big science" approach is needed to build the testbeds required for research to advance.
Fixing the Internet and increasing computer science funding are both excellent goals. On the other hand, one may ask: is the academic community in a position to create a new Internet? Will the planned testbeds permit realistic experimentation? Is building testbeds the best use of scarce funds?
I'm both skeptical and supportive. I'm skeptical that academics will have a significant impact on future Internet evolution: there are far greater forces at work now--for example, Amazon and Google. And I don't see how network research testbeds can support the applications needed to identify the *real* problems. (I am reminded of "Touch's Law," coined by Joe Touch: the lowest level at which a testbed allows experimentation is also the highest level at which experimentation can occur.) On the other hand, independent of these issues, the program will get smart faculty and students building things on a large scale, and that has to produce innovation. So while I'm not sure about the end, I applaud the means.
Genies traditionally offer three wishes, but I have only one for GENI: that the program engage the users, researchers, and engineers from the eScience community. I sense that "eScience" is viewed by networking researchers as irrelevant to the demands of the "real Internet." But looking back over time, eScience applications have arguably been one of the biggest drivers of computing and networking innovation: think early Internet work, the Web, clustering, graphics, and many other things. (And Grid.) eScientists are doing amazing things, struggling with difficult issues, and pioneering new approaches to building and operating community infrastructures--and their future goals are yet more ambitious.
I believe that getting eScience applications engaged could ultimately be seen as transformative for the program.