1. The Rule of Three - Sheth and Sisodia. Why every market will be dominated by just three (generalist) players and a bunch of specialists. Ought to be wonderful, but it's a hard read. Full of facts and examples, it's a Guardian (SJ) book. There's an interesting theory trying to get out, but it's incredibly well-hidden. Frustrating.
2. Feynman's lectures in Physics - Vol III. What a joy to read! But the thought/text ratio is so high that it's not a fast read. I guess a six month course over at Caltech.
3. The Economics of Regulation - Kahn. Another kind of worthy book full of dense but interesting ideas. But it's huge, and getting through it is such a slow process.
4. Chronicles of Amber - Roger Zelazny. First read when I was a teen, and I had forgotten how good it is. For when I'm too tired to think.
I installed BOINC today - the Berkeley Open Infrastructure for Network Computing. It is not a user friendly system: there is real confusing stuff about account keys (no longer used, apparently but you still get error messages). Configuring preferences for your computer is obscure and hard to find, and there is little guidance as to what values to set, what they mean and what might be the consequences.
In the end, I rolled over, and just let BOINC do whatever it wanted with my computer. It seems restrained so far, running the Einstein@Home gravitational wave detector and the Rosetta protein folding experiment. But guys, I was persistent: you can do better than this!