Thursday EWT took us to the Blue Lagoon in the morning, and then in the afternoon we got back to hacking. The Blue Lagoon is a hot spring about 45 minutes away. It was a nice break in the mid-week. After a quick lunch, we got on the bus and headed back into town for some more Python hacking.

I'm not sure if there were very few completed projects on Thursday, or if they just didn't make it into the list of projects that people were working on. Perhaps that means people were primarily just continuing work from the other days. Probably it's because Steve didn't have the list up on the "big board" during the review after lunch.

I know in our case, it was simply continued work. Richard, Georg, and I got to the point where we could run some tests of the new exception classes. The initial test results are that the single largest pybench performance drop has turned into the single largest gain. However, additional work will probably slow it down some amount.

Work continued on the string routines, when Fredrik wasn't stirring things up on the Python mailing lists. Continued improvements there are looking great.

Overall performance improvements, according to my tests, were fairly modest at around 1% according to pybench as compared with 2.4.3. Remember that we started the week 10% slower than 2.4.3, so this is nothing shabby.

Work continues. I can give you a quick peek ahead, big gains will be reported tomorrow. :-)
(Post Reply)