digga VS rfcs

This one issue may clear things up, seems like my config is a little outdated: https://github.com/divnix/digga/pull/385

I'm new to the Nix world, but so far I've come across Divnix's Digga, Numtide's DevShell, and Misterio77's nix-starter-configs.

There are config templates / configuration helper libraries that try to make this easier, for example digga/devos.

https://nixos.wiki/wiki/Flakes, especially the “see also” section. If you’re looking to use for NixOS config across multiple hosts, digga (see the repo for example template) is pretty nice for encapsulating a lot of boilerplate.

The digga library, while being more complex to use than other solutions here, got a pretty elegant solution for it merged a few weeks ago. Still some cracks that are getting smoothed over, but it seems to work.

I like the concept digga/devos uses (unfortunately their stuff kind of is an overengineered incomprehensible mess): They use: - modules: for modules like in nixpkgs (i.e. stuff that defines options and generates configuration based on that options; are included into every host) - profiles: concrete configuration, can be included to host definitions - suites: sets of profiles (so you can for example have a desktop suite with all your profiles with "desktop" configuration options and apply that to all your desktop computers)

For the system, I like the devos template:

https://github.com/divnix/devos

The idea of flakes is how you define inputs, and you define the system (and packages, and shell etc.) in the outputs using the inputs. The inputs are git repos which point to other flakes. You can mix and match these as much as you want (see the devos repo for examples) and when you build the derivation, it generates a lockfile for exact commits in that point in time what were used in the given inputs.

You commit the lockfile and in the other systems where you pull your config from the repo, it uses exactly those commits and installs the same versions as you did in your other systems.

This was quite annoying and hard to do before flakes. Now it's easy.

The problem what people face with building their system as a flake is combining the packages so you can point to `jq` from the unstable nixos and firefox from the stable train. I think this aspect needs better documentation so it wouldn't be so damn hard to learn (believe me, I know). Luckily there are projects like devos that give a nice template for people to play with (with documentation!)

Another use for flakes is to create a development shell for your repo, an example what I did a while ago:

https://github.com/pimeys/nix-prisma-example

Either have `nix-direnv` installed, enter the directory and say `direnv allow`, or just `nix develop` and it will gather, compile and install the correct versions of packages to your shell. Updating the packages? Call `nix flake update` in the directory, commit the lockfile and everybody else gets the new versions to their shell.

Yes, I put the repository in ~/nix. My repository is based on devos, but I am thinking of switching to a different setup, because I don't want to depend on a framework which can be an issue in updating.

I no longer use the official way since I have switched to flakes. I am currently using a devos-based config, which is a boilerplate that depends on a Nix toolchain, but I plan on rewriting the config with flake-utils-plus. You probably can install home-manager using deploy-rs. See the following comment:

> (after eelco ignored the PR for quite a while, also!)

Clicking that link takes us to a PR that was opened on 2024-02-02. The initial response from the Nix author comes 7 minutes later. Puck has multiple back and forths with other members Github, but her next interaction with the Nix author comes the next day on 2024-02-03. This is also the first time in the conversation where she "reminds him ... to even read her PR message". There's a second interaction later that same day during which she does similar, but it's worth noting this is pointing to a different message and appears to be less a "reminder to read" and more re-iterating what they feel is their argument against the Nix author's own arguments. Puck then continues to have back and forth with other commenters but as of today, there has been no further comments from the Nix author after 2024-02-03, and no further comments from Puck after 2024-02-08.

This hardly to my mind qualifies either as "having to remind him multiple times to even read her PR message at all" or "after eelco ignored the PR for quite a while, also!" So as I said it's a fairly weak claim, and feels more like a "bastard eating crackers" reaction to the PR than an actual showing of poor behavior.

As for the "Meson example", I didn't ignore it. As I stated in my comment, I had at that point read two of the referenced discussions in detail, and thus commented on them. I didn't comment in the "Meson example" for the simple reason that I hadn't read it.

I have read it now, and equally find it confusing.

1) The claim in the letter is that the proposal has "passed RFC, for five years", yet the RFC itself only appears to have been opened 2022-08-24. It's been a while since grade school for me, and I'll admit COVID has warped all our sense of time, but I'm pretty sure 2022 is not 5 years ago.

2) The first completed working implementation of the change doesn't appear to have been done until 2023-01-18 (https://github.com/NixOS/rfcs/pull/132#issuecomment-13874661...). Again this is much less than 5 years old.

3) On 2023-03-20, the author of the PR for this change states:

> the RFC has made it past most of the early stages and the current goal is to achieve parity with the current buildsystem before replacing it.

(https://github.com/NixOS/rfcs/pull/132#issuecomment-14768433...)

Again, this doesn't seem to fit at all with the claim that the proposal has "passed RFC, for five years"

4) On 2023-11-01, the Nix author themselves asks for updates on the RFC implementation, an action which doesn't seem congruent with someone who is willy nilly single handedly blocking things and being a disruption to the process. And the author of the PR states:

>the main block is actually a lack of free time for the main devs!

(https://github.com/NixOS/rfcs/pull/132#issuecomment-17890770...)

This doesn't seem to point to evidence that the Nix author is single handedly holding up this process.

5) On 2024-03-21 the PR author notes:

> currently working on adding support to build nix-perl, waiting for assistance

(https://github.com/NixOS/rfcs/pull/132#issuecomment-20135356...)

Not to sound like a broken record, but if the issue isn't finished as of a few weeks ago, it can hardly be considered to be held up by the Nix author for 5 years.

I agree that one of the links in the open letter is to a comment on a PR from 2019, which is indeed 5 year ago, and does indeed contain the Nix author commenting that they are skeptical of the change because "he doesn't know meson but knows his own build system". But given that there's an entire wealth of history on the topic since then, including progress on the feature that appears completely unobstructed by the Nix author and an open PR that is a mere 3 weeks old for a current implementation, I find myself again unconvinced of this rampant bad behavior on the part of the Nix author. And I reiterate again that these complaints are very weak and don't do much to support the open letter at best, and act as contrary evidence at worst.

Again there might be other context to be had that is missing, but if one is going to write a massive "open letter" complaining about bad behavior, I expect the links in that letter to point to actual bad behavior, and or provide the relevant context necessary to show how what appears to be normal dissent is a passive aggressive continuation of obstruction. I have to assume the links one provides in an open letter is their strongest evidence, and if this is all the authors have... I am unconvinced.

Nix with dynamic derivations (RFC92) could potentially beat this curse.

https://github.com/NixOS/rfcs/blob/master/rfcs/0092-plan-dyn...

See: A plan to stabilize the new CLI and Flakes incrementally https://github.com/NixOS/rfcs/pull/136

NixOS recently introduced "problem" infrastructure to deal with such problems more gracefully and explicitly:

https://github.com/NixOS/rfcs/blob/master/rfcs/0127-issues-w...

For some more context: Flawed as they are, Flakes solve a large number of problems Nix experiences without them. This is why I, and presumably many others, use them even in their current experimental state.

An RFC was recently accepted to commit to forming a plan towards stabilization of Flakes: https://github.com/NixOS/rfcs/pull/136

Personally, I don't believe there won't be any breaking changes, but I also believe that the stabilization of Flakes is still a ways away and hope that there will be a reasonable migration path.

NixOS needs to merge https://github.com/NixOS/rfcs/pull/133 to solve the issue