Opportunity is missed by most people because it is dressed in overalls and looks like work. -- Thomas Edison

You betcha. An opportunity for a big project has knocked. It looks like lots of work. More details once I figure out how to take it on.

I've mentioned Copernic Summarizer, which distills web pages, before. I liked it enough to upgrade when my free trial expired recently. Copernic made me a bundled offer I couldn't refuse and persuaded me to get Agent Professional as well. Agent is the kind of product I've shrugged my shoulders at many times, but it turns out to be extremely useful.

Superficially, it's just a metasearch engine, combining results from a whole bunch of search sites. However, it includes all sorts of nifty tools to analyze the results for duplicates and broken links, extract key concepts, filter by date, add your own notes, and on and on. You can save searches for further use, or schedule them to run at intervals you specify. Very cool.

As a materials scientist, it's sort of humbling to realize that the average spider knows more about engineering materials than I do. Spider silk is the strongest natural fiber known, with properties surpassing many manmade fibers. The February issue of Applied Physics A includes a special section on the materials science of silk.

One of the great joys of conferences in general, and Materials Research Society meetings in particular, is the chance to find out what's going on in areas outside my main focus in integrated circuits. Like, say, molecular electronics.

While I'm highly skeptical about the prospects for commercial molecular electronic devices, people in that area are doing fascinating basic research. I was especially impressed by their beautiful experimental technique. When you're trying to measure the behavior of individual molecules, even an AFM is something of a blunt instrument. But you can use micromachined structures to make atomically sharp gold wire contacts. You can use nanowires as very small optical probes. You can treat all of your measurements as statistical samples, and repeat them often enough to see real data emerge from measurement noise.

The idea of measurement statistics especially reminded me of an observation someone made about semiconductor manufacturing metrology: if your measurements are fast, you can make a lot of them. Conversely, as noise increases, you have to make a lot of measurements, so they had better be fast.