Ask HN: Why don’t you use Linux as your daily driver OS?

You made their point though. Food being Vegan or Halal does not matter to people who don’t care about that. You don’t need to sell to people who already buy in. The software has to stand on its own merits against the competition. A FOSS license is very very rarely going to be the thing that pushes a user to drop Photoshop for GIMP

Specific applications.

Inability to use it for work.

Lack of a deep selection of polished, well-supported apps.

Then you underestimate the number of people who care about FOSS. Consider then a niche like 'power users', 99% of the user base doesn't need to care or use a feature for it to be an important aspect and have a trickle-down effect in who uses it and thus who develops it. People into FOSS commonly contribute to the development of tools they use.

The key thing is choice. You don't have to use the software if you don't see the value in it, likewise with features. Just like you don't have to choose gluten-free or vegan foods, but quite commonly, lots of very popular dishes are vegan, halal, gluten-free, keto-friendly etc., and there is a not insignificant chance that you already consume it in some way and enjoy it, and likewise a strong chance your every day life is fuelled by FOSS, even if it is just part of a server stack on a website you are visiting.

All the software someone uses does not need to be FOSS, however certain pieces of software might be more easily trusted in security and privacy if they are FOSS. Hence F-Droid's apps for example.

Likewise, I don't care if a game isn't FOSS, but I will care about my note-taking software if I am writing down vulnerabilities I've found for responsible disclosure.

With Adobe having cloud connectivity, and unknown telemetry, if I want to view PDFs that may contain sensitive information that some naïve intern compiled, I won't be using Acrobat, that's for sure.

I contribute to several projects because I actively use the tools, and yes, the license matters.

1.) I don't have a Linux laptop

Given my previous experiences, I am never again running Linux on a laptop that is not meant to be a Linux laptop. I have owned a System76 but now I own a Windows laptop (former died and I had to buy the latter in a pinch). I spend my hard earned money for every component on the computer, so I won't accept that something doesn't work just so I can say Linux is installed. Should I be in the market for a new laptop, I would then consider a System76 or the like. The learning experience of installing Arch was great, but it's just one of those things I'm doing once for educational purposes and never doing again. It's not something I want as a hobby. Next time, as soon as I open my new laptop, I just want to install the software I will be using, not fixing hardware related issues. With my System76, it was the other way around; I installed a windows boot for work and it didn't work as well as the Linux boot. So morally of the story. as obvious as it is to me now, is to run the OS your laptop was supposed to run for the best experience.

2.) I'm jaded from the last time I installed a distro.

I did what all the cool kids were doing, stopped using a mainstream distro, and installed Arch. It's working and it's there. I don't boot it anymore. The amount of work it took to get everything working was non-trivial to say the least. After that certain things broke with updates. Several other things never worked or had its quirks.

2.) The community

After spending so much time installing my last distro, I suddenly realized that the amount of time I was spending on the Linux desktop qualified as it being a hobby. And when I realized that and then took a look at the community, I finally saw it; they're all hobbyist. Linux is their hobby. They like to distro hop and WM hop, rice their configurations, then post on r/Unixporn. So it became clear that it was less about productivity and awesome tooling, which was the original intent, and more about ricing your desktop which obviously Linux is way better at.

4.) I run a different OS now, Emacs

First I tried running Emacs on windows and certain things like Magit were painfully slow. I am now comfortably running Emacs with WSL + X410. Using Emacs as a front end to my OS makes the underlying "backend OS" (I'm making that up now) less important. My reasons for switching to Linux years ago was because I saw this cool thing called i3wm that could let me control my desktop the way I was controlling my vim editor; with my keyboard. However, after my Linux laptop died and I switched to a Windows one in a pinch, I needed to find a replacement to that environment on the Windows side. Naturally started with WSL, then had a fairly productive tmux + fish + vim configuration, and now I'm on Emacs.

Granted, Emacs runs better on my Arch boot compared to WSL + X410. Still, I'm pretty happy with this new setup.

5.) I develop windows software for work

This might be a bit anticlimactic but had I not chosen a job that developers windows software, I'm sure I would probably still be on Linux. Because I need to keep going back to Windows for work, whether a VM or booted, it has forced me to figure out how to get the tooling I loved to use on the Linux side working on the Windows side. This led me to reason 4 and emacs.

I know many of you will suggest to split work and personal, but the circumstances of my work/life make it way more practical to just have everything together, in particular because the usual "work computer" constraints don't apply to me.

-----------------------------

Would I switch back to Linux? To fix the above issues, here's what I would do;

1.) Buy a Linux laptop

System76, Dell Dev edition, Slimbook, Tuxedo, are all good options.

2.) Use a mainstream distro

I just want my OS to work so I can start Emacs

3.) Change jobs (highly unlikely)

4.) Stop listening to the community

If I were to install a Linux distro for my father again I would probably use NixOS, because he won't be installing packages, and then just maintaining the config file, automatic upgrades and killing older generations just works. Staying with stable and possibly having to patch the config sligtly (to fix upgrades) every 6m.

The learning curve is higher once you wanna start building software that isn't packaged already, that's when you'll have to learn to write nix expressions.

I've been planning to package FRRouting for NixOS, It'd be the perfect OS for a router, atomic forward and backwards rolls. (Except for user data, but that could be solved with ZFS snapshots, NixOS makes it VERY easy to use ZOL).

Nix is super unique, you can set up your own Hydra (the Nix build bot), run your own package cache, fork and maintain your own stuff. It's really more like working with code than maintaining a system.

Another great thing is that with Nix, once a problem is solved once, It's solved forever (Since literally everything defining your system is under version control).

I love it, but i wish to learn the Nix language proper and contribute back.