U of T's Aran Donohue has created Academic Hacker News, and response has been very good: as he says in his blog, it seems to have real traction. There goes whatever productivity I had left... Meanwhile, Aran's asked for good models of what an MSc thesis should look like. I pointed him at a couple of ACM Distinguished Dissertation winners (Ebeling's All the Right Moves and Gibson's Redundant Disk Arrays are my favorites), but (a) those are PhD-sized, and (b) they're almost as old as he is. What would you recommend someone model their Master's thesis on?