Recurse Center, 2014-07-17

Kyle and I did a demo of our spectrum analyzer/visualizer during the presenations. It was fun to work on, though we mostly just followed a tutorial on the web, and made use of a bunch of libraries.
I spent the night in HackerSchool.
I was cleaning up the code in the tutorial we were trying to follow.
Also, cleaned up the install.sh used by lightshowpi project to not do all the ugly sudo setups, and use a Python virtualenv and install into that.
Refactored the ugly looking music part of the tutorial into a smaller script with only the functionality that we were going to use.
We hit an interesting bug that would light up all the LEDs on the strip, once in a while. I didn’t notice it during the night, because I had a “decay” factor (the max factor by which the height of the columns should get reduced between successive updates) was 0.9, but when the decay factor was reduced to a lower value, it would happen quite often. Also, we didn’t see this happening before. So, we thought the bug was in the code I had written, when I should have been sleeping. After reading and debugging my code for a bit, with Kyle and Sean, I thought it was something hardware related that we were doing wrong. But, it turns out that a library we were using lit-up the whole strip, when start and end values were both zero, instead of not lighting up anything!
Also, we hit another off-by one error, just before the demo. I didn’t do the math for splitting the strip into columns too carefully, and we hadn’t noticed the off-by one error until we taped up the strip into different columns on a pillar for demo.
There were some cool presentations by others. I’ll update the list on Monday, since I currently don’t remember them all! Looking at the list of names on the registration sheet would help!
The cleaned up code is here