[–] 9 points 1 month ago (4 children)

It's on line 3: nixpkgs.url = "github:NixOS/nixpkgs/nixos-unstable".

Change it to github:NixOS/nixpkgs/nixos-26.05 for the latest nixos stable (which is not rolling, meaning you'll have to update the URL every now and then)

  • source
  • [–] 3 points 2 months ago

    Really cool but also significantly limited by QEMU-in-WASM performance. I know Nixpkgs has some WASM cross-compilation capabilities, wonder if it can be hacked further to enable something like this running more "natively"

  • source
  • [–] 7 points 2 months ago (1 child)

    I've been using it for several years now. It did take me about a year to learn it before switching though.

    My previous server was the same install for about a decade old and it took weeks to unpick everything and migrate, the random edit to config years earlier that's apparently critical to something i don't even remember until it breaks or the cron running on a service account i forgot exists.

    I love that it's all in one place now with git history

  • source
  • parent
  • context
  • [–] 5 points 2 months ago (4 children)

    NixOS is wonderful. At the same time it's pretty annoying how often I need to test a workflow at my job and it just doesn't work because X or Y downloaded a prebuilt binary that can't dynamically load libraries. The solution is usually simple, but the fact that it's recurring sucks.

  • source
  • [–] 13 points 2 months ago (3 children)

    NixOS has ruined all other operating systems for me.

    It’s made actually running a secure system so much easier, I love that when I define a new service I can create and manage service accounts and permissions very clearly and reliably. It makes it hard to screw up user permissions, which was something I hated dealing with in Linux before.

  • source
  • [–] 4 points 2 months ago

    Very nice! One of the issues of using Nix has been the ad-hoc methods for pinning versions, such as rust-overlay. (Although rust-overlay is great for toolchain customization which might make it worth keeping around.)

    Even with some ad-hoc pieces I'm happier using Nix than using completely unrelated tools for each ecosystem, like Rustup and Nvm. If we could have a more unified Nix solution that would be even better.

    It could certainly make my old Python versions test matrix easier to maintain.

  • source
  • [–] 5 points 2 months ago

    That's the master branch and, as you noticed, it was added after the 26.05 release (that's why I didn't find it when looking through home manager options). So it's likely that after that module is finished and released, niri-flake won't be needed.

    (To be pedantic, niri-flake was never "needed", it just provides useful options/functions so you don't have to write them yourself. And it could be the case that it will still provide some things that the home manager module won't, though I can't think of any examples)

  • source
  • parent
  • context
  • [–] [S] 2 points 2 months ago (1 child)

    It seems like wayland.windowManager.niri.settings is defined in home-manager/modules/services/window-managers/niri.nix. Is this not all you need to configure Niri through home manager, or do you need something more? Actually, looking closer at the repo, the earliest commit to niri.nix is actually from 28/07/2026, so it might just be that the home manager module is so recent that information on the internet hasn't caught up yet.

    Thank you.

  • source
  • parent
  • context
  • [–] 10 points 2 months ago (2 children)

    I think the main reason is that niri is not included in home manager by default. So that repo adds the ability to configure niri through home manager. nixpkgs only provides the package to run the niri binary. Niri-flake adds handy config options, creates the config file, etc. Also I think there were some issues with the "official" niri package, which that flake solved? I dont remember the details.

  • source
  • [–] 7 points 3 months ago*

    To be clear, they tried to keep the team alive. Originally it was 4 people. Due to the amount of pressure "from below" (as in, lots of requests to decide on how to do something in Nixpkgs, with whatever decision the team makes stirring up controversy) and the lack of communication "from above" (as in, the difficulty in figuring out the boundaries of their authority with the steering committee) two people burned out quite soon. The remaining two issued a public call for more volunteers and even reached out to people directly, but only one person was interested and AFAIU they did not end up joining the team.

    Now, finally, both of the remaining members burned out, and with noone stepping up after repeated calls for new volunteers, the team has no other choice but to disband. No reason to keep a team that's literally empty.

    As for their open letter, I think it was well-written and not combative or performative in any way. They pretty much just announce they are disbanding, explain the reasons why without putting any blame on any individuals, and thank everyone for their help.

  • source
  • parent
  • context
  • [–] 25 points 3 months ago

    just seems like extremely poor communication and more "open letter" bullshit that's really a nothing burger. Seems like one or two people said "we're disbanding!" while based on some threads I've read the remaining team was like "yeah...we really didn't want to"

    There is no remaining team. The two people who are resigning here are the entire team (originally it was four members, but two had already resigned previously). What's the point of having a team without any members?

    just privately quietly resign

    How are you supposed to "quietly and privately" resign from a community project that versions team membership and responsibility in Git? Genuinely, what are you imagining that is lower profile than this and not wildly irresponsible?

  • source
  • parent
  • context
  • [–] 1 point 3 months ago

    I feel like the effect of the social media zeitgeist has turned a lot of people, especially those born in this century, into permanent performers. That's not to say that their actions are purely performative, rather that to them if it doesn't happen in view of the entire internet then it may as well have not happened at all.

  • source
  • parent
  • context
  • view more: ‹ prev next ›