Linux

Knowing this stuff is fine but make sure to keep your goals in mind. If the idea is to get a job, figuring out how Bluetooth works isn't going to get you anywhere. You need to move in the direction the wider industry is moving. That direction is running containers in kubernetes.

If you can stand up a kube cluster, write a Prometheus exporter in go, scale pods based on those metrics, and auto resize workloads' resource requests, then you should be able to find a job without much trouble.. These are the things ops people are expected to do in 2023.

EDIT: The CNCF is a great resource for modern tooling.

BeyondLFS. I did both and ran it as my daily on a K6-3 for a couple years.

Some things to check on ur lfs:

• bluetooth audio
• screenshare (with media audio)
• mic quality (sometimes there's a lot of static sound if not tuned properly)
• try running a game
• try running other resource intensive software like training an ml model or running 3d cad software
• try writing to files that you're not supposed to be able to (without using sudo), to test that you gave the proper read/write/execute access to system files.

I've done LFS myself. After the core build, I moved on to BLFS to actually create something close to an everyday OS. I also tried to index the packages. But after a while, I got tired of all the work and moved on to other distro. I finally settled with Gentoo, with which I can customize the package compilation with way less effort.

LFS is a great tool for education & practice, but we couldn't expect it to be productive as there are more effective & efficient tools out there that can do the things you do (and probably more) with way less time and effort.

One thing one would need to keep in mind is the great deal of effort to maintain the build on top of all the effort creating the build in the first place. If done by the book alone, it'd be like Sisyphus pushing a boulder uphill.

You can make automation tools for managing packages (or use automated LFS, part of the LFS project), but chances are the ones already out there can do better (e.g. apt, dnf, pacman, you name it). In fact, any tool you'd likely think of would've probably been created and tested.

With that said, you still can learn a great deal. Maybe it would give you all the knowledge and skill to do better, might even come up with something new.

Learn ZFS. I guess it isn't included in typical tutorials, but if you spend a few days with practising on it, and also read some of the 'deeper' stuff about it, then afterwards you feel like you can do magic :-)

After you do LFS try making a minimal bootable system (on a stick) and see what's the bare minimum you can put on it to get a working console. Hint: init can be any executable or script, it doesn't have to be complicated.

You can also add useful tools like parted and make your own little mini-rescue distro that you can use when something goes wrong with your main system.

The Linux Foundation hosts brilliant courses on OS virtualization in Linux - after that there is a pretty clear path at a cloud administrator career.