Very interesting. This doesn’t look at all like what I thought it would look like.
In fact, it looks like the code that reads the motion data from the chip is working just fine.
That leaves a couple of possibilities…
One is that we’re getting some floating-point errors in the fusor code. I think we would be able to detect that by using the “dumpfusor” command.
Another possibility is some error in the gesture-detecting code. If that’s the case, this problem could start happening 35.5 or 71 minutes after boot.