On 01/06/2010 08:47 AM, geni wrote: I could be wrong about what process was being used before; I had just assumed based on external observation. I agree that Duke Nukem Forever is an example of that process failing, but you can find failure examples for any scheduling approach. In the Duke Nukem case, assuming the recent Wired article was correct, their goal was to be radically better than every game existing, which meant moving the goalposts every time they saw something new and shiny. That would have broken any scheduling model. A lot of open source projects work pretty happily on the "it's done when it's done", so I don't think there's anything fundamentally wrong with that approach. To me, it's more a question of which approach best fits the circumstances. William