| 8 Sep 2021 |
@gytis-ivaskevicius:matrix.org | I have went over it few times by now, but yeah... Cant say that it seems tempting | 08:24:42 |
David Arnold (blaggacao) | I think it cultivates an "outside-in" perspective and treats nix as cattle, not as pet. I like that. | 11:58:29 |
David Arnold (blaggacao) | * I think it cultivates an "outside-in" perspective and treats nix as cattle, not as pet. I have to admit that I pretty much like that. | 11:59:15 |
David Arnold (blaggacao) | It seems somewhat affordable to the average nix-averse developer. And might be _the_way to implement a task runner in nix without self-debyingt (toml files). | 12:00:14 |
David Arnold (blaggacao) | * It seems somewhat affordable to the average nix-averse developer. And might be _the_way to implement a task runner in nix without self-debying (toml files). | 12:00:19 |
David Arnold (blaggacao) | * It seems somewhat affordable to the average nix-averse developer. And might be _the_ way to implement a task runner in nix without self-debying (toml files). | 12:00:30 |
tomberek | It (makes) did make a few interesting choices, though as-is - Iām hesitant. Learning from the design and the experience is valuable. Iād like to see flakes used under the hood. (I think it would improve performance and reusability of the library it made). | 12:37:21 |
David Arnold (blaggacao) | I have the same feeling, though I saw that in an earlier iteration, they seemed to have had a wrapper around flakes, so I don't know how difficult that would be to reimplement. For many, not embracing flakes might be a show stopper. Otoh, it looks like they did not prioritize this implementation detail over an overall better user experience (no "special" install). | 13:56:59 |
David Arnold (blaggacao) | I wary, though, what that means for their use of fetchers - which - as I understand it, will eventually be denied within a flake's output function. | 13:57:55 |
David Arnold (blaggacao) | For those who remember the nix build issue I reported on a few days ago: https://github.com/NixOS/nix/issues/5227 | 14:50:37 |
| 9 Sep 2021 |
@princemachiavelli:matrix.org | Just updated my flake.lock... I'm getting this error when trying to build error: attribute 'agenixRoot' missing Anyone have this issue? | 00:31:19 |
@timdeh:matrix.org | hmm can't say I've seen that yet. Is this a new option in agenix perhaps? | 00:50:02 |
@princemachiavelli:matrix.org | I don't think so since ryantm/agenix hasn't had any commit since Aug 1st. I'm going to go back and update each flake input individually. | 00:51:33 |
@princemachiavelli:matrix.org | It seems to break when I update the nixos input. The trace is pretty much all in the nixpkgs/lib code. Seems like something in nixos-unstable is broken. | 01:17:48 |
@princemachiavelli:matrix.org | Looks like there was a change to the activation scripts around removing users/groups in PR 136605 witch was merged today. This is what fails to evaluate: ${textClosureMap id (withHeadlines) (attrNames withHeadlines)}. Not sure if this is a Nix issue or a Devos issue. | 01:25:07 |
| 10 Sep 2021 |
| @aynish:sealight.xyz joined the room. | 23:47:24 |
| 11 Sep 2021 |
| andi- joined the room. | 03:29:07 |
@aynish:sealight.xyz | hello fellow nix nerds, ive been using nixOS for a year or so now, and getting into flakes. switched over to hlissners dotfiles, but wanted to replicate my own config, so im in the process of switching to using devos, nice to meet yall šš½ | 15:45:53 |
| yusdacra changed their profile picture. | 16:32:35 |
David Arnold (blaggacao) | Hi aynish! Welcome! Make sure to regard digga as your tool and devos as your template. | 22:31:29 |
| 13 Sep 2021 |
| Maksim joined the room. | 08:36:01 |
Maksim | Hi there :-) I'm really curious in exploring digga and devos but hence the amount of abstractions I don't know where to start. Currently I'm using nix to manage my hm config in a wsl context and share the config with a nixos installation on a server, but as it's really brittle (e.G. I have to symlink my home.nix each time I update the config) I'd like to create something more consistent. | 08:39:14 |
Maksim | what it is this flk vs the digga tool? | 08:39:40 |
@aynish:sealight.xyz | flk is just a wrapper around nixos-rebuild (iirc) and digga is a standard api for handling flakes. since you'll be combining new derivations (one you create) with ones you consume (that others created) digga makes it easy / standard to only add the former to your outputs, so you're not messing up downstream flakes | 10:53:11 |
David Arnold (blaggacao) | Please also note that flk has been superseded by divnix/bud. And alltogether recently fluidattacks/makes popped up the screen, which is fishing structurally in a similar pond ar bud (just not specifically made to interact with a system configuration). | 13:50:31 |
David Arnold (blaggacao) | * Please also note that `flk` has been superseded by `divnix/bud`. And alltogether recently `fluidattacks/makes` popped up the screen, which is fishing structurally in a similar pond as `bud`, just not specifically made to interact with a system configuration. | 13:51:09 |
David Arnold (blaggacao) | * Please also note that `flk` has been superseded by `divnix/bud`. And alltogether recently `fluidattacks/makes` popped up the screen, which is fishing structurally in a similar pond as `bud`, namely a task runner with a CLI store-front, just not specifically made to interact with a system configuration. | 13:51:48 |
| 15 Sep 2021 |
yusdacra | Is makes as modular as `bud`? What I love about `bud` is that you can create commands for it using anything you want and share it easily. Maybe `makes` could be used together with `bud`? | 06:50:10 |
David Arnold (blaggacao) | To me it looks like makes is a generalized version of bud. Makes is a devops task runner, bud is a devos task runner. | 13:03:37 |
David Arnold (blaggacao) | * To me it looks like `makes` is a generalized version of `bud`. Makes is a devops task runner, bud is a devos task runner. | 13:03:49 |