i followed a “install any linux distro in 5 minutes” guide and after just 14 hours i finally have an OS that is running properly

as a long time hater i must say its very nice and i like it a lot, especially considering microsoft is hellbent on making windows the worst OS as quickly as possible. i will say, im a big nerd and this stuff comes to me pretty easily but i still think hoping normies will adopt this is kinda insane

anywho, its the year of linux or so i keep hearing

also if anyone knows why my keyboard keeps disconnecting every 30 minutes or so, i would very much like to know why. its an anne pro 2 that i am using with a wired connection

  • dead [he/him]@hexbear.net
    link
    fedilink
    English
    arrow-up
    10
    ·
    1 day ago

    I used this method to install the nvidia drivers on Debian. I have the driver version 580.105 installed, newer drivers are causing bugs for me. The repository contains pinning packages to stay on a certain driver version.

    wget https://developer.download.nvidia.com/compute/cuda/repos/debian12/x86_64/cuda-keyring_1.1-1_all.deb
    sudo dpkg -i cuda-keyring_1.1-1_all.deb
    sudo apt update
    sudo apt install nvidia-driver-pinning-580.105.0
    sudo apt update
    sudo apt install linux-headers-amd64
    sudo apt install nvidia-driver
    

    https://docs.nvidia.com/datacenter/tesla/driver-installation-guide/debian.html

    • GoebbelsDeezNuts [any]@hexbear.netOP
      link
      fedilink
      English
      arrow-up
      7
      ·
      edit-2
      1 day ago

      I knew I should have asked here first when I thought about it yesterday. I looked at Debians documentation and about a billion threads and GitHub repos and the only thing I didn’t think to check was nvidias documentation. I feel stupid.

      Also thank you if I decide to try it again I’ll do this. I’m kinda just trying to enjoy it at the moment after fighting with it all day yesterday

      • For a LONG time, the conventional wisdom was to shun the packages provided by Nvidia (as well as AMD in the proprietary FGLRX days) in favor of distro packages. These hardware companies used to ship distro-agnostic installers which would install the drivers and tweak the config files once, but they did not integrate with the package manager at all, and would inevitably break the next time you update your OS. The only way to use these drivers reliably was for distro maintainers to re-package them.

        Nowadays, AMD support is automatic (built into the kernel / Mesa3D), and Nvidia actually hosts their own package repositories instead of just a self-destructing installer. You should still prefer a distro package if its available, but the knee-jerk “never download the Nvidia driver from Nvidia” advice is not as true as it used to be.

      • dead [he/him]@hexbear.net
        link
        fedilink
        English
        arrow-up
        4
        ·
        1 day ago

        The documentation is confusing. The official debian wiki does have a link to the page. That repository is maintained by Nvidia for datacenter usage but it works for desktop drivers as well.

        The page below has a tool for how to add the Nvidia repo on different distros. It also has the option to set up your own local repo. Do not use the install script named “runfile”, it could brick your distro install. Only install the nvidia driver through the distro package manager.

        https://developer.nvidia.com/cuda-downloads