I love that you have a very specific and arguably moral crime in mind, and it inspired this post.

Did you grep that log file for 'amdgpu'?

I wonder if the error is related to this: https://github.com/NixOS/nixpkgs/issues/229108

I'm still using x11 on my system. Maybe try that and see if it works?

No, I haven't seen anything like that. That's odd.

I've had those errors on my system for years. I never thought that they were NixOS specific. I just assumed something to do with a buggy firmware:

Enabled 4 GPEs in block 00 to 1F
ACPI Error: Aborting method \_SB.PCI0.GPP2.PTXH.RHUB.POT3._PLD due to previous error (AE_AML_UNINITIALIZED_ELEMENT) (20240322/psparse-529)
[x~20]
ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-ff])

I don't notice any ill-effects from them, so it may be a red herring. I have a:

$ < /sys/devices/virtual/dmi/id/board_name
ROG STRIX B450-F GAMING

with a 5900X.

I don't usually see as many prints as you have there, but it's quite a few, and the number seems to vary (grow?) over time. I keep meaning to investigate it, but haven't got around to it.

I think you should keep looking in your logs for other problems. If you can share the full log I'd be happy to take a look.

I'm not afraid of scissors, but I'm very afraid of my victorinox knife.

Am I just failing to use that site properly, or is it missing a ton of stuff in 'replays' that was available live?

I feel like the CBC had a better version of this thing 12 years ago.

I guess at the 2028 Olympics they'll be jumping on the AI bandwagon.

You're suggesting people not be able to run software in kernel mode on their own systems.

I would never run kernel mode anti-cheat, but going down this road will lead to hardware attestation and the end of open computing for anything with online services.

Why would there be one answer to this? I'd probably use all the available levels depending on the situation, in the same way I'd use --word-diff or -b in git when I need help understanding a complex change.

The original error actually makes it sound like there's a partition on hda that's bigger than hda itself.

It probably becomes CPU limited with those other compression algorithms.

You could use something like atop to find the bottleneck.

I was thinking about this last time I drove an ev (ioniq 5). It will really decelerate quite hard when you lift off, and it's configurable by the driver.

I don't think they need to do it with an accelerometer, but if the regeneration system is applying more braking force than it would take to turn on the light with the brake pedal, it should turn the light on.

Either that or they should require the brake pedal to be used beyond that point.

Edit: actually it just occurred to me that it might be no worse than downshifting in a normal car. Maybe it's not a big deal.