I would say the German system works pretty well and is well accepted but the outrage when they ruled that the lid on plastic bottles needs to be attached to the bottle was beyond infantile

Haha touché. Maybe I should have added that he is a conspiracy theorist (of the Reichsbürger kind)

I had an argument with my uncle once. In Germany, there is “Pfand” on bottles and soda cans. You pay a little extra when you buy one and when you return it to the “Pfandautomat” you get back the extra amount. He was arguing that this is somehow indoctrination by the state to condition us to follow rules. I have avoided conversations with him since

It definitely depends on the circumstances. With 60 GHz radar e.g. you get quite a good distance resolution and can detect e.g. breathing rate really well (from the torso movement during breathing) and things like how many people are in a room, etc. But its always very dependent on the environment, your settings, subjects, noise, whatever. That’s why I said its typically not practical. By using dedicated devices perhaps, and most of these kinds of news are about people who use dedicated devices, but that’s like putting a camera in your home. When you have to abuse an existing communication channel, probably not so realistic.

My topic was fall detection (as in elderly people falling) specifically without using cameras or wearables. The idea was to take the CSI (basically what you see in the image) and just stuff it into some machine learning model to get a prediction as to whether someone fell in a given time frame, so I was trying to classify the signature of the falling “activity”. From my literature survey, this has been done successfully with CSI. But as with a lot of research, it typically lacked practicality. Much of my work was implementing the firmware, data recording, processing, and so on. I also had to record a ton of falls (ouch) and label them. I ended up throwing away the CSI approach though, because of the noise reasons I mentioned. That was simply a deal breaker. I went with FMCW radar instead (and it worked pretty good).

Interesting. I didn’t actually read into BFI details, thank you

to be fair, maybe. To pass FCC/CE regulation regarding EMC, it has to adhere to strict limits at 2.4 GHz (but I could also imagine for microwave ovens specifically that the allowed emissions are higher than for other devices, because 2.4 GHz is just the band it operates in. But idk, I didnt read the standard for those). But it does not mean that it may not radiate anything in that band.

Anyways, my observation was that it did interfere and the microwave was definitely closed. But also it was not 10m distance to the microwave, more like 2m, so relatively close. WiFi receivers are quite sensitive to be able to work with low received powers. So just a little emission is sufficient to interfere. You are probably not disturbing the communication itself, because OFDM is quite robust, but it certainly destroyed my use case (which operated on the whole CSI).

And there is definitely some stuff leaking, e.g. through radiated emissions on the wiring (the power line). But it is certainly not cooking anything. That’s also what the regulation makes sure of.

I think the main advantage with the wifi-based approaches is that they are usually used in a relatively static/calm indoor environment with a stable channel response and your motions are disturbing that, compared to a quickly changing outdoor environment (e.g. a city) where it would be much harder to distinguish individuals. Also, you are typically closer to the access points, making the power/SNR higher. Regarding mobile communication though, the trend is towards higher frequencies and smaller cell sizes which also give greater spatial resolution (and higher power) and some funky near-field effects can be used to get beam forming on crack: https://arxiv.org/abs/2407.10147 So perhaps it could work even better, wouldn’t be surprised

I mentioned this in a another reply, but you just need a microwave oven basically :)

That’s using CSI though. The article said the researches specifically did not utilize CSI.

But regarding CSI: I evaluated that as a small part of my Master’s thesis and it worked pretty OK for motion detection but not for classifying other activities, at least not on a SISO link. For more complex stuff you would need both a MIMO access point (router) and device (e.g. phone). Also, you need to constantly transmit messages to get up-to-date CSI, which is not great for power consumption as well as cluttering the communication channel. There are some other constraints, especially regarding noise. E.g. I managed to completely destroy the CSI spectrogram by turning on a microwave oven. There is 802.11bf in development, which is supposed to standardize this, because currently using CSI is pretty much a “hack”, as it is not intended for sensing. Once this is widely adopted, I would start being worried, but not right now.

This is from my thesis:

For me it was getting rid of the chair and just sitting on the floor + floor desk. I am very naturally changing positions often and for some reason check my posture a lot more than when sitting in the chair. There are some communities online related to this. But it does take some getting used to. Same with sleeping on the floor, that was even more important for my back. But I cannot do it at work and I guess it requires that your legs and knees are still OK.

PIDs for can be very fun. And essentially its just 5 or so lines of code, which is something I wish they told us in uni (instead, it was mostly theory, as in the meme). I recently built a kiln which goes to 1000-something °C with a PID controller and I just set the parameters by vibes, not even some formal method. And it just works. So here is my resource: The (bit messy) controller code for my oven The code is obviously a bit more than those few lines, but I just wanted to say that the implementation is very simple, which I would have liked to know when I started out with this.

Edit: just found a bug after looking at the code again haha, so thank you :D

I really noticed it when instead of thinking how to solve X, my mind started phrasing a prompt to ask how to solve X instead, does that make sense? I found this to be a dangerous, almost evil thing, and I’m sure it is the same with my coworkers, or they just don’t like to admit it. I still do this sometimes but am giving my best to unlearn it. And the crazy thing is, I did not even use it that much, only very occasionally, similar to what you mentioned you do. I do not wish to know how cooked the brains of “vibe coders” are by now…

Thanks :) Yeah, I didn’t show the ugly parts perhaps, haha. Especially the libadwaita apps (like Nautilus). I used to rice a lot more but don’t have so much time/patience for it now, and I think the current solution will stay for a while. Previously, I would have everything in one color scheme, i.e. bars, terminal, apps, editor, … but it is a bunch of work to maintain. So previously it would look something like this:

At a standup meeting, CEO asked whether we use AI and I was the only one who said that I don’t use it at all, or very rarely, starting a little discussion. Overall, their position is somewhat moderate. They do fall for the hype a lot (especially with the recent Claude stuff) but it did not seem that it was a requirement for us to use it. But they were curious why I do not embrace this so much and I said that I can feel myself getting more and more stupid when using these tools, due to the mental offloading. This seemed to resonate a bit with the others, at least I could feel that my coworkers in the round got my point, despite remaining silent.

Coworker recently came to my desk and jokingly asked why I was typing out code by hand when I could ask Claude to generate it for me, but there was also a bit of seriousness to it, so I cringed a lot

As I said, I personally wouldn’t do that in my free time, especially not every single day… because I’m too lazy haha

When you can only commute with a car, yeah it’s dumb. However, I have a daily commute of 1 hour with my bicycle and it’s a great way to get some exercise. In this regard it’s forcing me to move my body, which I otherwise probably wouldn’t in my free time. Gym of life.

Unfortunately, it is the only way my SO got paint tool SAI 2 with pen pressure working under Linux…

Plus, your source files would also have to actually be in that format. As an example, Spotify apparently supports up to 24-bit/44.1 kHz FLAC only

The observation is that most USB-C to 3.5mm (“aux jack”) audio dongles are supporting high quality audio but only support the USB Full Speed (FS) data rate which is not enough for the high quality audio. This causes problems.