I see often people say that the distro you are using doesn’t matter. One can turn any distro into another. And I do not agree with that. If that was true, why do we even have so many distributions? I always said, if distros don’t matter…

  • … why distro hop?
  • … why don’t you use Ubuntu then?
  • … why don’t you recommend Archlinux to a newcomer?
  • … why don’t you use Kali Linux as a server?
  • … why don’t you use Batocera or SteamOS as your daily driver?
  • … why do you trust a community distro more than a corporate distro? (or vice versa)

I don’t think that distros only matter to newcomers. Maybe it matters for experienced users even more.

  • IratePirate@feddit.org
    39·
    8 hours ago

    Just like your “opponents” are over-generalising, you’re deliberately picking the most extreme examples to make your argument. (Batocera as a daily driver - you know that’s what Hanna Montana Linux is for!)

    My Linux axioms are: for most new users…

    1. choice of DE is most noticeable and decides whether they like their initial experience.
    2. choice of base distro family does matter a lot in the long run (Debian-based vs. Arch-based vs. Redhat-based); if you stay inside the same family (e.g. Pop!OS vs. Ubuntu vs. Zorin vs. Mint), choice matters a lot less (and DE is most impactful, c.f. point one).
    3. choosing a distro with specialised security hardening (immutable systems, Nix, Qubes, Bazzite) does matter; most of these will make new users unhappy or even question their sanity.

    Where you are right: yes, the choices embedded within these three axioms do matter a lot and are noticeable, so it is helpful to have an experienced user recommend a distro to you when starting out.

    Where the “distro don’t matter” people are right: there are a lot less choices to be made than meets the eye. Effectively, it can be boiled down to three.

    • sudoer777@lemmy.mlEnglish
      2·
      7 hours ago

      choosing a distro with specialised security hardening (immutable systems, Nix, Qubes, Bazzite) does matter; most of these will make new users unhappy or even question their sanity.

      Out of curiosity, as someone who’s never used Bazzite/other uBlue/SilverBlue/etc, what makes it difficult for new users? I definitely agree with Nix and Qubes though (and SecureBlue to some extent).

      • IratePirate@feddit.org
        9·
        6 hours ago

        Basically, anything that isn’t packaged as a flatpak needs to be installed from the CLI using distrobox containers, which will go over the heads of the majority of new users.

        • Fecundpossum@lemmy.worldEnglish
          121·
          6 hours ago

          I can’t tell you how many times I’ve had to bonk the immutable Stans over the head with this logic, but it never works. They install bazzite, play their game of choice on steam and occasionally use their web browser and “ALL NEW USERS MUST USE BAZZITE ITS SO EASY”

          Any distro is easy when you use it like a fucking Sony PlayStation.

            • EchoDelta_9@programming.dev
              1·
              4 hours ago

              FWIW, uBlue has been brewing for almost three years now for their CLI stuff: see this issue tracker and this blogpost from Bluefin’s creator.

              The distrobox workflow overall has mostly been superseded by better alternatives[1]. Though, for completeness’ sake, openSUSE’s atomic offering continues to heavily rely on Distrobox. But, in their defense, I think their atomic offerings are simply better[2] suited for it.

              1. There’s sysext with its (WIP) manager, Brew Tap to tap into homebrew casks and some peeps even use coldbrew. And last, but definitely not least, nix support has improved over the years. And if you just want to use dnf, RakuOS’ innovative hybrid design allows just that; an image-based core you can’t touch (like the other ‘immutables’), but dnf works and is applied through a persistent overlay. ↩︎

              2. Fedora’s container images are tied to its major release versions. Hence, every 7-13 months you’re required to set them up from scratch if you’d like to continue using them 😅. Even if this process can be streamlined, it’s IMO very cumbersome regardless. In openSUSE’s case, the containers are based on Tumbleweed. Which, has a rolling release cadence. Hence, it was meant to be used indefinitely. ↩︎

              • IratePirate@feddit.org
                1·
                3 hours ago

                Thank you very much for the detailed and well-sourced write-up! I’ve saved it for later when I get to drill down on this.

                It kind of proves OP’s point though: distros do come with a lot of idiosyncrasies of “how things are done around these parts”.

      • EchoDelta_9@programming.dev
        1·
        7 hours ago

        Not the one you asked, but I think the answer lies in the bold part:

        most of these will make new users unhappy or even question their sanity.

        For example, I can’t imagine any of the uBlue projects causing major difficulties. Though, edge cases do exist; adding kernel mods can still be a bitch, even if there are efforts to improve this.

    • thingsiplay@lemmy.mlOP
      15·
      8 hours ago

      There is another point, which makes this discussion very variable. Also the choice matters or does not matter a lot, depending on the person, expectations and what is being done. This is probably the biggest reason why we don’t agree on simplifications like these. And BTW, just because the examples I gave are extreme does not make them wrong in any way. They are just easy for illustrating my points I’m making.

      If someone is coming from Windows, does not care much about trust and just want something that runs a browser, doesn’t care about community or technicalities, then yes it does not matter if the person chooses Ubuntu or Mint. On the other hand, if someone doesn’t like corporations, has strong opinions about standards and is a developer, then the choice suddenly matters a lot more.