Heroic also has an option to move games to a different directory so you can move them between drives easily after they’re already installed.

I ran Debian as my main OS for years around 2005-2010 or so. If you run the testing branch it gets you pretty new stuff. Just don’t run the testing branch for a few months after they do a stable release as that’s when all the big breaking changes take place. Also, Trixie is still relatively new and has a new enough kernel to make most GPU stuff work well. If you need a newer kernel it’s not too difficult to build a new one and use that so you can get fresher amdgpu modules, for example.

All my jobs back to 2002 until now I have had full authority to run any OS I like on my work computer. I’ve run nothing but Linux on all my work machines and I have convinced many coworkers to do the same.

And use CEPH as your filesystem.

You’re quite a bit behind on BIOS version so I’d update that first and then run memtest86+ which you can install from the arch repo and it will show up in your grub menu.

BIOS update can be found at https://www.gigabyte.com/Motherboard/B450M-DS3H-rev-1x/support#dl

I have the same printer and it works great on Linux.

btop shows it and a ton of other info too.

My favorite Linus moment ever.

I run SELinux on tons of servers at work. We taught our Oracle consultants how to use it. Some software vendors get mad at us because we require it and we always figure out how to make it work and it isn’t all that bad to work with once you’re used to it

Sounds like you have an issue with your PATH for the user you’re sshing as. What does ‘/usr/bin/echo $PATH’ output when run via ssh to your server?

I loved the glorious 3.5.x days. What a fantastic DE it was then. I compiled 3.5.0 from source when it was released because it was going to take the Fedora guys too long to package.

I can’t believe they’re up to 40. I remember installing Fedora Core 1 like it was yesterday. Yum (and now dnf) has come a long way. It used to have to individually retrieve metadata files for every available package, rather than using a single compressed index of all the packages available in the repository you were using. It made just getting to the stage where dependencies were calculated take forever.