Show's over (success!)

Thanks to MS covering for us last night, and an excellent working session today, the show at Adler went off very well. JF and SS are great banterers.

This weekend's the first time I've ever been convinced that seeing stuff on a planetarium dome actually shows you more of the data. And yes, it is nice to see stuff we've made go up on a dome. And yes, I got to control it, which was fun. The navigational system SkySkan's Digital Sky has more degrees of freedom than Partiview has, which is very nice, but I took a while to learn that I had to press both mouse buttons (instead of just the left one) to stop motion.

Everyone seems to think this is quite a big deal. There were lots of congratulations and back-slapping going around afterwards. And, given the audience (some milionaires, at least one billionaire, the University president, the Adler president, etc) I suppose it is, but I'm so jaded and fed up with infovis and with myself (I should know OpenGL/VRML/VTK/Swing/Director by now, but know nota) that it doesn't feel like much. Just like another show. And tomorrow I've got an interview with a company for machine learning stuff - let's see how that goes.

Just thinking - if I had to stay in academia, I'd like two years to concentrate on Information Visualization. I would like to build a system for visualizing SVMs and kernel methods, including semisupervised and active learning. And I'd like to build a system to create 3d models of data using Partiview to begin with.