The above repeats an objection that I've seen raised a couple of times.
The .dev TLD might invoke the feeling that Nix is only meant for developers
I agree with that.
NixOS can also be used by non-developers.
It can, but should that be a concern? I argue: No, we should very deliberately keep the focus on developers.
At the moment, and for the foreseeable future, we don't have the resources to do anything else well. The only way I see to get them is by making the overall system more reliable and therefore indispensable for those who (can) use it as it is right now and will invest time or money to improve it. And unless we were explicitly selling end-user convenience, increasing accessibility will therefore have to be created from some surplus, which I currently also don't see appearing.
This statement by a (back then) new user is also relevant:
And it makes me feel very unwelcome and gives an impression that I'm not a target audience at all. However, after using NixOS for a while I really dig it and I really feel that I am. And I want less people to be intimidated by its website.
To be clear, I want NixOS to be useful and the community and technology appear welcoming to anyone, but let's be realistic about what can be done with what we have. Making things easy to use for people with very limited time is surprisingly expensive. Doing that for a huge legacy system even more so. It doesn't preclude anyone from trying though, but I'd prefer not pretending it's somehow around the corner.
And then, what does it even mean to be a developer? We had a similar discussion among Nix maintainers more than a year ago, and while it didn't lead to concrete answers, I think it was quite informative. Arguably you don't have to be a professional or have a formal education, you just have to develop software, which today means manipulating code. All this requires is a particular mind set, Bash, cUrl, and then Nix to get all the other tools needed for that, and a lot of time to learn new things. Nixpkgs and NixOS, in my opinion, are great not just because they give you ready-made artifacts - that's merely a side effect of caching, specifically the public cache - but because it allows and encourages customisation of the software that produces them.
By the way, all the documentation effort is really about is reducing the cost of learning those new things; by uncovering, structuring, and sequencing the knowledge a bit, also so it's easier for learners to hold realistic expectations of these costs.
All that said, actionable alternatives:
Remove this statement without replacement
Put it under a section for considerations raised, and explain why it's not an issue
https://github.com/nix-rfc-canonical-domain/rfcs/blob/d2535a22c9c638aa8c40e842dc414d6065159b45/rfcs/1000-canonical-domain.md?plain=1#L51
The above repeats an objection that I've seen raised a couple of times.
I agree with that.
It can, but should that be a concern? I argue: No, we should very deliberately keep the focus on developers.
At the moment, and for the foreseeable future, we don't have the resources to do anything else well. The only way I see to get them is by making the overall system more reliable and therefore indispensable for those who (can) use it as it is right now and will invest time or money to improve it. And unless we were explicitly selling end-user convenience, increasing accessibility will therefore have to be created from some surplus, which I currently also don't see appearing.
The marketing team clearly states that their target audience are software developers, and @garbas has driven that course since at least 2020. The documentation team, from the very first day, was supposed to align with marketing efforts.
This statement by a (back then) new user is also relevant:
To be clear, I want NixOS to be useful and the community and technology appear welcoming to anyone, but let's be realistic about what can be done with what we have. Making things easy to use for people with very limited time is surprisingly expensive. Doing that for a huge legacy system even more so. It doesn't preclude anyone from trying though, but I'd prefer not pretending it's somehow around the corner.
And then, what does it even mean to be a developer? We had a similar discussion among Nix maintainers more than a year ago, and while it didn't lead to concrete answers, I think it was quite informative. Arguably you don't have to be a professional or have a formal education, you just have to develop software, which today means manipulating code. All this requires is a particular mind set, Bash, cUrl, and then Nix to get all the other tools needed for that, and a lot of time to learn new things. Nixpkgs and NixOS, in my opinion, are great not just because they give you ready-made artifacts - that's merely a side effect of caching, specifically the public cache - but because it allows and encourages customisation of the software that produces them.
By the way, all the documentation effort is really about is reducing the cost of learning those new things; by uncovering, structuring, and sequencing the knowledge a bit, also so it's easier for learners to hold realistic expectations of these costs.
All that said, actionable alternatives: