I'm partially peeved at Puzzle Pirates. And not just for alliterative purposes. I enjoy the little games, but I discovered after spending my earnings on a new pair of pants (for my avatar) that only paying customers get to wear nice pants. I knew that some items and games were pay-only but I'd screwed up my clothing order. :( I may be done with it for a while; I think it contributed to screwing up my back last night. I should stick to games with "pause" features.
I'm reading "Selected Papers on Analysis of Algorithms" by Don Knuth. I can't honestly say I'm following all the math. (Some of it is recognizable from my "concrete math" course, though.) It's sort of interesting seeing what sort of problems were interesting over the past decades, though. Like algorithms for merge sort using tapes: is that a dead field, or is there somewhere some tenured professor still churning out theorems...?
I think I was motivated to read the book (I've owned it for a while) by a comment made at DDB's party, probably by sethb, about probabilistic proofs. The key idea in such proofs is to show that if Prob(x has property Y) > 0, then there is some x with property Y... even if we have no idea what x might be. Or you can work it the other way around and show that P(x has property Y) < 1 and so there is some x with property not-Y. I think it would be fun (for very geeky definitions of fun) to look for some of these predicted but not-known objects.