Monday, May 31, 2010

Delay and such

Alright, time to really, really, really get back on the blogging horse. This time for real or something but most folks should probably know better. Must be too content as things are largely good :)

Anyway, still owe a longer post on the previous topic with regards to how academia and research views security versus the real world. Various events interspersed since that post have really only reinforced that view but I'll withhold judgment until the process makes it way to completion. Once it works its way to completion, I promise a scathing rant.

On a secondary note, I've been doing a bit more pondering with regards to the principles of admission control and in particular systems in general. I am hoping but perhaps being naively optimistic to spend some serious time trying to sketch out a paper for HotNets this summer on the topic. The basic crux of the paper is pretty simple, we are wrongly focusing on minutia when in reality, we should really be focusing on dynamics. In my humble opinion, dynamics are what we should really be thinking about when it comes to scale. How do we keep the dynamics under control, i.e. avoid rapid change? My intuition is that we can use this as a guiding principle for system design in that things should be limited to change within a certain bound, thus limiting the emergent dynamics that can occur that we typically fail to understand.

Not sure exactly how I will go about crafting this paper but HotNets tends to align nicely with it. Worst case is that I write a nice screed that gets rejected but is ultimately quite cathartic. The public / private firewall paper seemed to go quite well along those lines despite its rocky road to publication. The paper still brings a smile to my face even almost a year after publication.