(no subject)
Yesterday included a Microsoft presentation on the next version of Visual Studio. I think the sample application (something about real estate) was a poor fit for the audience, but the tools themselves were impressive. The editor can do all sorts of crazy things to your source code based on the AST, but when I asked the presenter about accessing that programmatically, he said it won't be possible until version after next. He also showed me that the compile and build tools for CIL targets are actually distributed with the runtime, and VS is just a pretty face, which really surprised me.
The wildcard round was a bit of an anticlimax, as it resulted in a final room containing the top four seeds. (One could argue that this means the system is working as expected.) The finals set seemed a bit easier than the previous rounds, but it wasn't obvious to all the competitors that the set of strings to count in the hard problem was a regular language.
Microsoft had given out these magnetic stick-and-ball sets, and last night I persuaded a bunch of people to play Zendo with them. It was interesting because these are very natural for building graphs, and I tend to look at even the standard Icehouse pieces as being in a graph, although I think modeling geometric relationships was the intent of the game designers. But playing with a group of programmers means that graph-oriented rules are natural and quickly solved.
Off to lunch with what remains of the TCers.
The wildcard round was a bit of an anticlimax, as it resulted in a final room containing the top four seeds. (One could argue that this means the system is working as expected.) The finals set seemed a bit easier than the previous rounds, but it wasn't obvious to all the competitors that the set of strings to count in the hard problem was a regular language.
Microsoft had given out these magnetic stick-and-ball sets, and last night I persuaded a bunch of people to play Zendo with them. It was interesting because these are very natural for building graphs, and I tend to look at even the standard Icehouse pieces as being in a graph, although I think modeling geometric relationships was the intent of the game designers. But playing with a group of programmers means that graph-oriented rules are natural and quickly solved.
Off to lunch with what remains of the TCers.