I was in the audience when this talk was presented. They really didn't say much about the actual way they were generating or mutating their inputs, just the sys

I was in the audience when this talk was presented. They really didn't say much about the actual way they were generating or mutating their inputs, just the system they hacked together to distribute the fuzzing runs, and from a security point of view, that's way less interesting. Charlie's talk on fuzzing was much better, imho. (Charlie's slides: http://bit.ly/dt3dkb http://bit.ly/dt3dkb (10meg PowerPoint file.)