Nix Documentation | 414 Members | |
| Discussion about documentation improvements around the Nix ecosystem | 84 Servers |
| Sender | Message | Time |
|---|---|---|
| 18 Jun 2023 | ||
| This seems to agree with the position that it’s not Nixlang that people struggle with, but the basic concepts | 18:08:28 | |
| Maybe? | 18:08:58 | |
| I agree with this view | 18:10:19 | |
| For the longest time I couldn't understand what flakes were and what utility it's provided over using a nix shell. | 18:11:17 | |
| Redacted or Malformed Event | 18:12:03 | |
| The most illuminating piece of exposition I saw was the Reddit post that actually made me gain a good understanding of it: https://www.reddit.com/r/NixOS/comments/13ye3lg/what_are_the_advantages_of_using_flakes/ | 18:14:16 | |
| I think a lot of very basic concepts in Nix such as overlays, channels, derivations, binary caches, etc are often overlooked when giving motivating examples of its use. | 18:15:41 | |
| And that is the biggest impediment to Nix adoption | 18:16:11 | |
| should we wait for roberth to look at nixos ¬docbook or just send it and go on to nixpkgs? | 18:25:51 | |
| 20:48:32 | ||
| DominicMills: Those seem like implementation details that are uninteresting as motivating examples though? Except for binary caches. | 21:45:52 | |
| Why would a potential user care about derivations, for example? | 21:46:20 | |
| I think I don't even do and I'm already sold on Nix. | 21:46:44 | |
| I'm a software developer and care about derivations. I believe many software developers would care about creating their own derivations for whatever projects they work on. This is especially true for complex polyglot projects that would otherwise have a number of moving parts without "packing" them into a derivation. Moreover, creating derivations for your own projects comes with a number of benefits such as package deployment and distribution, dependency management and versioning and isolation for testing purposes, which would be useful for a large of subset of software development teams. I find it interesting that out of all topics I mentioned you picked derivations to criticise, since it is arguably has the most utility compared to everything else that I've listed. | 22:15:28 | |
| * I'm a software developer and I care about derivations. I believe many software developers would care about creating their own derivations for whatever projects they work on. This is especially true for complex polyglot projects that would otherwise have a number of moving parts without "packing" them into a derivation. Moreover, creating derivations for your own projects comes with a number of benefits such as package deployment and distribution, dependency management and versioning and isolation for testing purposes, which would be useful for a large of subset of software development teams. I find it interesting that out of all topics I mentioned you picked derivations to criticise, since it is arguably has the most utility compared to everything else that I've listed. | 23:29:43 | |
| * The most illuminating piece of exposition that I've seen was this Reddit post that gave me a good understanding of it: https://www.reddit.com/r/NixOS/comments/13ye3lg/what_are_the_advantages_of_using_flakes/ | 23:30:44 | |
| * For the longest time I couldn't understand what flakes were and what utility it provided over using a nix shell. | 23:30:49 | |
| 19 Jun 2023 | ||
toonn might be talking about raw derivations rather than builders like stdenv.mkDerivation which is what a developer is much more likely to use | 08:49:09 | |
| I like this explanation of the build process, would be nice if we could have something similar on nix.dev https://news.ycombinator.com/item?id=36388251 | 08:59:52 | |
| I'm talking about derivations, which are the product of something like a call to mkDerivation and contains the concrete information Nix uses to build something. I didn't know DominicMills was using the term for package and shell expressions. | 21:18:03 | |
| Just a problem of terminology. Which should definitely be a consideration for the docs though : ) | 21:19:46 | |
| don't the docs call everything derivations? especially the pills | 21:20:13 | |
| 20 Jun 2023 | ||
In reply to @pennae:matrix.eno.space Because everything is a derivation. Using a different word for derivations in Nix code vs derivations in the Nix store would probably be even more confusing. | 01:56:26 | |
| I created this post to solicit feedback on the tutorial series the working group is working on: https://discourse.nixos.org/t/2023-06-19-tutorial-series-call-for-feedback-1/29377 This is part of the new process we're trying out to be more productive and leverage experts in the community since most of us in the working group are beginners. | 02:53:12 | |
| 08:30:53 | ||
| Alex: I disagree. There's a small set of Nix expressions that result in derivations. They all end up calling the built-in `derivation` in the end AFAIK. | 10:33:53 | |
In reply to @toonn:matrix.orgBut almost all Nix code eventually leads to such a derivation. Setting up those derivations is usually the whole purpose of the code. | 10:38:14 | |
| I hope the documentation can be clear and helpful while still being accurate. | 12:35:14 | |
| nixos-render-docs probably needs a good round of refactoring after the dedocbookifications are done, the current mechanism it uses really doesn't work all that well with what we ended up. markdown-it has a syntax tree module that'd probably be a much better fit to how we handle data. 🤔 | 17:35:20 | |
| 21 Jun 2023 | ||
| pennae: you are working on nixdoc related dedocbookification now right? can you say what would be required for https://github.com/NixOS/nixpkgs/issues/126375 and when it would make sense to do it? I'd be interested to pick it up sometime again, but I figure it doesn't make sense while stuff is in flux | 11:44:46 | |