I need to start off by mentioning that even as a realist with a strong scientific background, marketing can be a powerful force not to be underestimated - especially concerning aspects that are unfamiliar to oneself.

Water. H₂O. Why do they teach people in primary school that water's chemical formula is H₂O? Only perfectly pure water in a vacuum might approach that composition. Normal water is so much more complicated that trying to understand it is like trying to understand how individual electrons move in a macroscopic object like a person.

Enough rambling. I recently relocated to a small town in Alberta, Canada. The water used to be very soft on the west coast where I came from (TDS = 19ppm). So water treatment was really not needed in any home I have lived in. The water was soft enough to not cause issues with soap, washing machines, refrigerators or boilers, and hard enough not to be corrosive. However it is a different story here on the other side of the Rockies. I measured my TDS with an electronic meter and the result was 260ppm. That is considered very hard water (depending on who you ask). It is enough that I could see buildup of limescale on some of my equipment such as the HVAC unit. The previous owner had a Premier AF-40K water softener installed, however it had no salt in it.

I have been falling victim to a classic mistake - trying to compare something to some kind of perceived "average" or "normal" based on a sample of a group. It is easy to see a couple of instances of some behaviour and then assume it should apply to your sample as well, and if it does not, feel somehow disappointed.

See, the thing is Toruk cannot play fetch. He will sometimes chase after something thrown, but he will never bring it back. Other cats play fetch, even fish and whales play fetch:

Some days....

Why oh why?

waldo@waldomp ~ $ sntp time.windows.com
+2.074701 +/- 0.024010 time.windows.com 13.86.101.172

If I manually set the time correctly, then re-run this command the time is accurate. The moment I enable NTP time sync with any NTP server macOS adjusts the time to have a +2.07 second offset. This does not happen on my M1 MacBook Air, only the Mac Pro. Why?

I am using a HomeKit scene to turn on my Hair Light as well as my ATEM Mini Pro and Camera with one click, as enabling my video in my new setup is quite an involved process. Unfortunately I also have to manually turn on the Elgato Key LED lights using the Elgato Control Center status menu. This is a pain. So I thought to myself that I can simplify things by adding a HomeKit automation that whenever the Hair Light turns on / off, to have the Elgato lights follow it. Turns out this was more tricky than I thought...

First I tried to use Apple Script to interact with the Elgato Control Center status icon. This did not work - I could not get it to drop down and simulate button clicks. Next step was to try and see if I could reverse engineer the protocol between the app and the Elgato lights, and talk to them directly. First step was to get their IP addresses. It did not show in my Unifi portal and I could not locate it in pfSense, so I decided to run this on my firewall:

tcpdump -i ix0.8 -vv -nnvvXSs 1514 port 5353 | grep -i Elgato