Where people landed is that burnout was real, but the deeper problem is governance shape, not package management viability. Several commenters point out that the team existed for less than a year, so this is closer to an aborted org experiment than an existential product event. At the same time, others tie it to a longer pattern of resignations, moderation blowups, and fights around the steering committee. The common complaint is not just "drama" in the abstract. It is that responsibility and authority are badly matched. The steering committee is seen as too interventionist to delegate cleanly, yet too diffuse and overloaded to make decisions quickly itself.
A second theme is scale. Nixpkgs grew far beyond the size it had when looser governance was enough. People cite sharply rising contribution volume and the sheer breadth of the package set as reasons "we survived 20 years without this" is the wrong comfort. The project now needs more review capacity, clearer ownership, and better contributor protection than it had when it was smaller.
The thread also split cleanly between users talking about the software and people talking about the community. On the software side, even skeptics concede
Nix solves real problems. It is especially valued in large multi-language environments, robotics, fleets, and reproducible developer setups. The repeated knock is usability. Docs are fragmented, packaging quality is uneven, review queues can go dark, and the model feels worth it mainly once your environment is complex enough to justify the upfront cost. A few people argue LLMs are making that trade much better by lowering the barrier to writing Nix expressions and configs.
The mood is frustration more than panic. People still trust the technical core and keep shipping with it. What they do not trust is that the current social and governance setup can retain the people doing the hardest coordinating work.
discourse.nixos.org