| 24 Sep 2021 |
 David Arnold (blaggacao) | * That interface sits really at the `nix` level I guess. You could just override the `checks` output. (With `//`, for example) | 19:43:02 |
| David Arnold (blaggacao) | * I don't understand the gist of the "`emacs`"-issue, though. Can you boil it down? | 19:43:52 |
 @gtrunsec:matrix.org | emacs-issue? it seems to be an anti-namespace issue? | 19:44:37 |
| David Arnold (blaggacao) | Yeah, I didn't get it 😁😐 | 19:44:52 |
| David Arnold (blaggacao) | Maybe this helps:
rakeLeaves constructs a nested attribute set according to its crawling rules.
importables is passed as specialArgs, either to nixos or to hm.
| 19:50:52 |
| David Arnold (blaggacao) | At least that might give a better sense where something is breaking. | 19:51:23 |
 aciceri | In reply to @blaggacao:matrix.org
That interface sits really at the nix level I guess. You could just override the checks output. (With //, for example) But can I override checks in the flake outputs only in a certain case? | 19:51:38 |
| David Arnold (blaggacao) | You'd have to make your own version of checks. You could assign the mkFlakes output to something and then manipulate it. | 19:52:38 |
| David Arnold (blaggacao) | let output = mkFlakes ... in ... | 19:53:45 |
| aciceri | I understood what you say but I want to override my checks.aarch64-linux only if the host machine (I mean the machine where I executed nix flake check) has binfmt.emulatedSystems = [ "aarch64-linux" ]; | 19:54:10 |
| aciceri | This way the flake wouldn't be pure I think | 19:54:38 |
| David Arnold (blaggacao) | Ah, yeah. | 19:54:45 |
| David Arnold (blaggacao) | You could just build certain checks by hand. | 19:55:16 |
| aciceri | Can I read the "currently running configuration" inside a Flake? | 19:55:29 |
| David Arnold (blaggacao) | nix build .#... | 19:55:30 |
| David Arnold (blaggacao) | (one by one) | 19:55:34 |
| aciceri | In reply to @blaggacao:matrix.org
nix build .#... Yes, I could make a script that reads nix flake show and then run a sequence of nix build according to the machine where that command is launched | 19:56:44 |
| aciceri | A little bit too hacky maybe | 19:56:54 |
| David Arnold (blaggacao) | It's ok. The problem is in the teaying design, I guess. | 19:57:13 |
| David Arnold (blaggacao) | * It's ok. The problem is in the teasting design, I guess. | 19:57:42 |
| David Arnold (blaggacao) | * It's ok. The problem is in the testing design, I guess. | 19:57:52 |
| aciceri | Mmm, I can't understand what you mean | 19:58:04 |
| aciceri | * yeah | 19:58:42 |
| aciceri | However for a similar task I think this would turn out useful. | 19:59:08 |
| aciceri | Probabily using this and jq it could be a one line | 19:59:58 |
| aciceri | David Arnold (blaggacao): However thank you again! Sorry for the many questions and the bad english | 20:00:49 |
| David Arnold (blaggacao) | I mean why wouldn't you test all the architectures that you declared as supported in an environment where you can do that. | 20:01:47 |
| David Arnold (blaggacao) | Maybe Github is not the right tool. But with a little hackery, it might still be useful. | 20:02:15 |
| David Arnold (blaggacao) | I don't mean to push back, just set the context. 😆😁 | 20:02:47 |
 Pacman99 | I got in a PR to do the externalModules -> modules rename. I'm not very familiar with git submodules, what is the process to update devos with API changes?. I added a commit to the submodule in digga, but is there some way to connect that to a PR in devos? | 21:20:37 |
