Christmas news

I've been away from the blog for several days and I miss it! So, here's my catch-up edition. First, the sad news: our faithful Canon Digital Rebel XT died last night. An error 99 led me to a nice site with debug info (see update #2). After a bit of debug (shooting with no CF card, no lens, and a recharged battery) still showed the error, I noticed that the sensor cleaning mode didn't open the shutter. Oops -- almost certainly, the shutter is dead. We've used it for 4 1/2 years and close to 100,000 photos, so we've definitely gotten our money's worth out of it! It's now a toy for Alex and Daniel, so they're having a great time with it.

I've been doing some fun reading. It sounds like Boeing is finally ready with their latest plane, a small, economic airliner. It's a stark contrast to the gigantic Airbus A380, a concept I never liked. I've always preferred smaller airports to giant hubs and hope that Boeing wins out with a regional jet instead of a hub-to-hub behemoth from Airbus.

Another fun thing I ran across: nuclear reactor drawings (here are hi-res versions).

I finally had time to play with Skype and the Logitech Quickcam E3500 we bought several months ago. The two didn't work together and, reading on Google, I saw lots of complaints that the two companies blame each other for the problems. In particular, the webcam microphone wouldn't work, or sometimes worked inconsistently. I installed the latest drivers and software on the laptop and desktop to try again. The laptop worked fine; the desktop didn't. More Googling finally yielded the answer: if the mic volume is too high (>90% or so), then the webcam mic doesn't output anything. Unchecking the "Automatically adjust microphone settings" checkbox then manually bring the volume down fixed it. Amazing. Vika had a nice conversation with her family. There's lots of snow on the ground in Almaty -- they showed us a webcam shot of it!

I made some exciting progress on the professional fronts: I finally figured out how to deal with a troublesome term in the limiting-case analysis I'm doing with Durga. I'd thought about looking at a Taylor expansions, but didn't see any nice simplifications from that. However, playing with Maple, expanding just the first three terms produces a very small error. Wow -- I'd been puzzled by that for a long time!

I'm still excited about ARM stuff; I think I'll probably buy the proto board and see if I can find time to play with it during the semester. Speaking of the semester, some thoughts and plans:

For Micro, I hope to improve the course in three areas:

1. Make tests before I begin covering the material. That's just planning, preparation, and discipline.
2. Revise the labs and homework to require students in later labs to get an I2C, SPI, and A/D peripheral up and running. My plan is to develop that in the Spring then implement it in the Fall.
3. Create an in-class demo robot / revised Intro to Robotics robot / recruiting platform / summer camp robot. I think the same robot can serve all these goals, or at least the same basic robot with various minor modifications. Again, I'm planning on developing this in the Spring.

We'll see what I actually get done. That's quite a bit of work! I'll definitely be asking for some help with all this.