| 1 Mar 2024 |
|  tomberek set a profile picture. | 15:21:36 |
| 2 Mar 2024 |
|  Qyriad joined the room. | 06:56:26 |
| Qyriad | 07:51:29 |
| 3 Mar 2024 |
|  Tim changed their display name from Tim Le Pés to Tim. | 22:49:42 |
| 4 Mar 2024 |
|  nim65s joined the room. | 12:02:12 |
| 6 Mar 2024 |
|  Sammy (It/Its) joined the room. | 19:01:49 |
| 7 Mar 2024 |
|  dyerat joined the room. | 08:49:53 |
| dyerat | Hello :)
Is there any reason I shouldn't move all of the config from within my configuration.nix into my flake.nix and then remove the configuration.nix file? | 09:07:51 |
 samasaur |
- modularization; 2) version control
| 09:08:37 |
| samasaur | for 1, let's say you have two nixos machines, hostA and hostB. then you could have a common.nix file that contains config that they both have, and then have a hostA.nix and a hostB.nix file that contain config unique to machines A and B respectively. then flake.nix could define the system configurations and import the corresponding files as modules | 09:10:18 |
| samasaur | for 2, keeping flake.nix and configuration.nix as separate files means that (usually) when changing the machine config, you only make changes to configuration.nix, and (usually) when restructuring config/adding new machines or other outputs, you only make changes to flake.nix. this means that when you look back in version control, both types of changes are not combined in the history of a single file | 09:12:34 |
| samasaur | but neither of these are requirements — if you would prefer to have everything in one file, you can do so | 09:12:55 |
| dyerat | for the nix.settings.experiemental-features = [ "nix-command" "flakes" ] line, would that always be required in the configuration.nix file / always requiring me to home one called exactly that | 09:14:45 |
| 8 Mar 2024 |
| nim65s | Hello !
I have project A which has nixpkgs + flake-utils as inputs. fine. Then project B which depend on A, so nixpkgs + flake-utils + A (whose nixpkgs & flake-utils inputs follows B inputs).
Then for project C which depends on B… I am supposed to write nixpkgs + flake-utils + A (whose nixpkgs + flake-utils inputs follows C inputs) + B (whose nixpkgs + flake-utils + A inputs follows C inputs), right ? It's going to be a lot of boilerplate fast, isn't it ? | 13:57:14 |
| nim65s | Maybe I should contribute A & B to nixpkgs, so that C only depends on nixpkgs 😅 | 14:05:25 |
| nim65s | * Maybe I should contribute A & B to nixpkgs, so that C only depends on nixpkgs + flake-utils 😅 | 14:05:37 |
|  @mtxyz:the-apothecary.club joined the room. | 14:06:47 |
|  @kyle:iteratee.net joined the room. | 20:00:30 |
| 9 Mar 2024 |
|  @qyriad:matrix.org left the room. | 00:36:30 |
 tengkuizdihar | In reply to @gsaurel:laas.fr
Hello !
I have project A which has nixpkgs + flake-utils as inputs. fine. Then project B which depend on A, so nixpkgs + flake-utils + A (whose nixpkgs & flake-utils inputs follows B inputs).
Then for project C which depends on B… I am supposed to write nixpkgs + flake-utils + A (whose nixpkgs + flake-utils inputs follows C inputs) + B (whose nixpkgs + flake-utils + A inputs follows C inputs), right ? It's going to be a lot of boilerplate fast, isn't it ? Wonder what the boilerplate looks like, but if this is a serious question, maybe you could post this on the forum. Its more discoverable that way | 02:51:03 |
