EDIT: As usual in some Linux forums, I got a bunch of marketing replies about all the cool ideas NixOS is supposed to bring as if I haven't read about them myself in LWN and other Linux sources I routinely check on.
At work, I'm maintaining a bunch of custom configured Linux images for our products. While building those images through Dockerfiles is pretty declarative, my pain starts with keeping the configuration of systems in the field up to date: half of the build instructions is duplicated in Ansible playbooks.
Conceptually, with Nix(OS), I could maintain a single, declarative configuration that I would use to create new images and simply push to existing machines an re-configure them.
But I don't see getting something like that production-ready in the near future. I'm currently thinking of only building base images with docker and doing all product-specific configuration through ansible.
There is nothing declarative about Dockerfiles, they are a list of imperative steps. Same for ansible playbooks, which are also often called declarative when they really aren't.
But on the abstraction level of configuring a Linux system this imperative nature can be completely hidden, as demonstrated by NixOS. E.g. you can set `services.uptime-kuma.enable = true` in a NixOS configuration and it will fetch the package, setup a systemd service for it, setup a service user, and enable and start the service, without you having to care about any kind of ordering or the specific steps themselves. You can do the same for other services, and it doesn't matter in which order they are declared.
Of course someone has to first build the abstraction, which is service-specific most of the time. Though, NixOS provides lower-level options to define these service abstractions declaratively as well (think: declaring that a user account is present, or that a systemd service is setup with some configuration), so most of the time you still don't reach the imperative base underlying NixOS.
I don't think that is the case. In five years of using NixOS I never needed to reach for any of them, and I don't even know what would be available to me. Probably something about activation scripts? Again, I can't think of anything you might want to do with regard to system configuration that can't be done declaratively on NixOS.
I've also introduced some light NixOS usage at work (3 hosts, one is an uptime-kuma instance, two are Forgejo Actions runners). For that I had to get some proprietary scanner software to run on it, which I could by just putting the extracted deb package in an emulated FHS environment and setting up a service for it, all declaratively.
Even for interfacing with legacy systems and unusual stacks I don't think you will need the escape hatch. Anything that is buildable on and above the abstraction-level of "ensure a file is present at some path with some content" should be doable declaratively, and that includes setting up an unusual software stack and running it in systemd services to communicate with some other legacy system or whatever.
The escape hatch is there to modify how NixOS itself behaves, and modifying that should only be necessary to extend NixOS' core functionality. A quick search revealed that impermanence (https://github.com/nix-community/impermanence) and in some cases sops-nix (https://github.com/Mic92/sops-nix) use it, but those fundamentally extend NixOS with ephemeral root storage support and secrets management, respectively.
Did it ever happen? Last I've checked flakes were still experimental.
99.5% percent of people don't use Docker because they have a passion for writing dockerfiles, or because they find the ideas behind Docker elegant. They use it as a tool to help build other things. Meanwhile, the Nix community keeps pushing out these ideas about how Nix _works_ ("reproducible builds!", "lazy evaluation!") but they don't seem to particularly care about making Nix easy to use for the majority of the population out there who may want to use Nix as tool and don't care about its technical merits.
On top, I think most projects would benefit from having more functional programming advocates in them. Nix suffers from the opposite problem: it would benefit from having _less_ passionate FP people in it. Nix the language is pretty inscrutable for lots of people.
If your threshold is based on the inscrutable language, I'm not sure that can be really solved by anyone in nix community at this point. You're just looking for a different solution, and that's fine. But nixos still solves real, practical issues for many people.
I'm just saying that, if the goal is mainstream adoption, the core team should acknowledge that people will use Nix as a means to an end, and not for the sake of using it as they do.
You're creating your own ideas of why people do things and presenting them as facts. Please don't, it's neither useful in the discussion nor nice.
> it's so easy to modify/remove/overwrite something
For some reason, out of the corner of my eye, I thought you were describing this as a feature of Nix because I missed the "mess up your install" part. It's easy to modify/remove/overwrite a package in Nix in a non-destructive way though. Of the few packages that have been broken for me, I feel empowered to pull them down and modify them to add needed functionality. And as I've gotten more confident, doing overlays or whatever they're called instead.
on nixos anything that you do has to be explicit in your config and thus it can be undone as needed.
So TL;DR -- because most programmers can't describe a use-case if their lives depended on it.
That's because "I think X sucks/is useless" isn't really a very useful comment in terms of the HN guidelines unless you're inviting discussion about the merits and drawbacks of the solution.
---
The other day, I upgraded my Ubuntu and it broke my display manager for some reason. I spent several hours debugging it. Given that investing in a niche technology where you won't get as much support as for e.g. Ubuntu (or even mac/win) has its own set of drawbacks, that isn't enough to make me switch to NixOS (problems like this really only ever happen every couple of years), but I do think that this isn't something that should normally happen. With NixOS, it would have been easy to rollback.