Anyway, did make positive progress on the compilers project, slowly and stumblefully.
Also made huge bounds on databases. I'm really happy that it's turned out to be somewhat substantial now. There was this one paper, which had two models, one of which was supposed to suck less than the other. And both of those models sucked on the wireless network traffic data we have, so we came up with another one, which was supposed to not suck, but actually it did, but in a different way, so we came up with yet a fourth model, which may in fact not suck. But the whole process of actually trying approaches and having them work-or-not, and trying new ideas is fun.
Played a brief game of go against dan. Lost pretty solidly. I should get back into actually playing go maybe over the summer.