bind mounts don’t usually have the problems that symlinks do. The only time I’ve had issues involve container systems like docker or flatpak.

For data like this from another filesystem I usually like to mount the entire volume somewhere private (like under /run) and then bind mount the parts I want to use into their desired locations (like /home/foo/Download, etc.)

I do this with a second ext4 drive that I use for performance sensitive storage with my primary btrfs system root. It works well, just be aware of edge cases involving containers (you may have to grant the container access to the original mount location under /run in addition to the bound path. This is definitely a gotcha when working with those paths with flatpak.)

Worst case scenario you could just bind mount the nonstandard steam directory into the usual place

Who needs an informed electorate anyway?

Boomers are just decrepit toddlers at this point anyway

Not understanding the difference between pre and post 9/11 politics

Why not just use something like Synergy so you can control both machines from the mouse+kb at your desktop? Just enable the software when you need to and you can move the mouse off the edge of the screen and onto the other machine as if it were a second monitor. That’s what I do with laptop + desktop setups. Get a small cheap Ethernet switch so you can plug both machines in.

I don’t even need passively cooled, solid state airjet cooling would be perfect for a <20W arm machine.

ASUS machines have solidly good Linux support these days thanks to the asus-linux community effort. Any of their newer machines (~2021-2023) will fulfill your ask. I’ve had a good experience with the 2021 g15 and the 2022 X16 - I’m using the X16 as a work laptop right now.

Qanon is like a clown car of unhinged conspiracy theories. Every time you think “Wow, that’s bananas, what kind of blood lead levels do these people have to fall for this?” they trot out something even more deranged/racist/fascist.

I’m onboard with 65 as the maximum age anyone can run for Congress but I don’t have a problem with people 65+ finishing their terms provided they’re actually competent. I’d like to see mandatory cognitive decline testing for anyone running for Congress, appointed to the SC or appointed to any high position in the executive branch.

It’s absolutely ridiculous that we’re allowing people with 5-7y remaining life expectancy to plan our future 20, 40 or 100y out - they just don’t have the skin in the game that someone in their 20s or 30s does.

On top of all of that I’d like to see vigorous corruption testing, SC justices and congresscreatures shouldn’t be bought and paid for the way they are now.

“Hey Internet, come do development on our product for free so we can monetize it. TIA”

Legal rights are not human rights. I suggest you go look up the definition of human rights, they’re a separate concept.

A country or state passing a law that makes it legal to punch clowns in the face on Tuesday doesn’t make that action a human right, it just means that country passes fucked up laws.

They’re also completely missing the point of distro kernel trees. Stable automatically selects patches from mainline (largely by keyword, and often without kernel developer feedback or involvement) and consequently has a massive amount of code churn and very little validation beyond shipping releases and waiting for regression reports. Distro trees are the buffer where actual testing happens before release. As a long term stable user it really isn’t suitable for end user or enterprise consumption unless you have your own in house validation process to test releases for regressions before deployment. Even running stable on client machines (desktops, laptops) leads to a bad time every few weeks when something sneaks in that breaks functionality.

I think Matomo is the popular ethical Google analytics alternative IIRC

Wait until you discover ventoy

Hardware accelerated JSON Markov chain operations when?

Time to train an LLM to format XML and hope for the best

The best I can do is an ML model running on an NPU that parses JSON in subtly wrong and impossible to debug ways