deathmtn

- Dropped off, picked up, dropped off.
- Did a couple hours of Client B work.
- Talked to a recruiter about a potential contract. May have fucked up the pricing talk.
- Examined the sample-reading part of the pd vocoder. I now understand it! It's using a line~ object to go through every sample in the incoming wav files. I thought it was doing something much more exotic. And the reason the logs looked so different was that I had let it run for a while, and being, a real-time system, something had gotten it messed up. When I stopped it after a single loop, the logs looked normal again. So, while I now understand more about how things work, I'm no closer to knowing what is messing up the C++ version. The post-sqrt/rqsrt multiplication, the kind of sqrt/rsqrt being used, and the block overlap and the next suspects to examine.
- Made dinner.
- Did some Client A work.
- Booked flights for February break. It wasn't that painful, though, of course still a little painful. It's nice that the ITA flight search matrix is still up. (Maybe for legal reasons?)

School's shut down tomorrow. It could go OK, but I was hoping to get more done tomorrow.

Aw crud. School's closed tomorrow.

The toilet was flushing OK but not great. I took the time to do 50 pumps of the plunger and now it perform excellently!

Nine-minute run, 12 pull-ups.

- Meeting with Client A. That was easier than expected.
- Got groceries.
- Went to the interview with the kid social group people. They seem like they do good work and are honest, but of course, it is expensive. Though we could probably ignore the guy's problems with reading cues and making friends, '80s-style, and it would work itself out eventually, this could make his life a lot easier. They said he's likely underconfident, though, and it may be true, but it made me feel bad.
- Got in two hours of work on Client B. My old goal was two hours a day on this, but for various reasons, this may not be attainable.
- Did some Client A work.
- Did some yoga. My right hip is total garbage.
- Poked for a while at the part of the pd vocoder that reads the incoming sources. It's surprisingly complex and reads samples every second. It uses a linear ramp, which I don't get. All I came away with is that I have to look at it more or replace it with something similar and see what effect that has.

12 pull-ups.

- Read about Q factors for oscillators. It's a thing I've used without actually knowing its definition. I don't get why first-order pass filters don't use it, though?
- Emailed a friend back.
- Did some Client A prep work.
- Did a few hours of Client B work.
- Got the guy to do some of his homework.
- Made dinner.

Not too bad for a day in which I also had the guy at home all day. All praise is due to Tinker Crate! He spent most of the day building a domino placement machine.

So, if you had a restaurant with an open garbage can and a fan blowing over the top of it, and you offered a discount to anyone willing to have some garbage air onto their dish, what discount amount would get at least half of the patrons to take the offer? This is an experiment I wish I could run.