| 10 Jul 2023 |
 figsoda | this would require patching the source code to use poetry-core instead of poetry | 01:23:51 |
| figsoda | it is technically possible with nix-init, but i think the patch should preferably be submitted to upstream | 01:25:23 |
| figsoda | maybe the best move is for nix-init to add poetry-core and print a warning stating that the source code needs to be patched | 01:25:55 |
 @blaggacao:matrix.org | I've two more that may be in scope:
- https://github.com/NixOS/nixpkgs/pull/242542#discussion_r1257646145
- https://github.com/NixOS/nixpkgs/pull/242542#discussion_r1257647353
| 02:39:32 |
| @blaggacao:matrix.org | TL;DR
(pythonOlder 3.9) if pyproject states itmeta.changelog if it's a github package (maybe)
| 02:40:28 |
| figsoda | added to https://github.com/nix-community/nix-init/issues/64 | 02:41:42 |
| figsoda | *
(pythonOlder 3.9) if pyproject states it
added to https://github.com/nix-community/nix-init/issues/64
| 02:41:54 |
| figsoda | I'm still a bit unsure about what to do with changelog in release notes, i've found myself manually adding those, but i couldn't find a good way to programmatically identify whether the release has meaningful release notes | 02:42:48 |
| figsoda | meaningful ~= anything more than the last commit message | 02:44:01 |
| @blaggacao:matrix.org | Yeah, agree. Impossible to give a good judgment with heuristics.
Are code comments in scope for nix-init?
| 02:47:40 |
| figsoda | I haven't explored with comments, but yes I would say so | 02:48:17 |
| @blaggacao:matrix.org | like so:
# TODO: uncomment if helpful
# changelog = "https://github.com/annotated-types/annotated-types/releases/tag/${src.rev}";;
| 02:48:28 |
| figsoda | anything that would help with packaging a program really | 02:48:40 |
| @blaggacao:matrix.org | I'm going right now crazy with nix-init... packaging essentially 7 python packages in parallel | 02:49:23 |
| figsoda | In reply to @blaggacao:matrix.org
like so:
# TODO: uncomment if helpful
# changelog = "https://github.com/annotated-types/annotated-types/releases/tag/${src.rev}";;
I will consider that, thanks | 02:49:41 |
| figsoda | i would still prefer heuristics if it works | 02:49:59 |
| figsoda | but maybe something like this could also work | 02:50:09 |
| @blaggacao:matrix.org | A small oped (ode) to nix-init | 11:41:46 |
| @blaggacao:matrix.org | * A small op-ed (ode) to nix-init 😎🙃 | 11:42:06 |
| @blaggacao:matrix.org | * A small [op-ed (ode)](https://discourse.nixos.org/t/disintermediate-documentation/30239?u=blaggacao) to nix-init 😎🙃 | 11:42:30 |
| @blaggacao:matrix.org | I just realized that pkgs-modules should probably expose programatic discoverability instructiona on theor interface definition for tools like nix-init to provide a runtime for them. | 12:25:42 |
| @blaggacao:matrix.org | I.e.
{
imports = [ (import ./pkgs-module-options.nix) ];
}
Should hav option annotations that let a tool like nix-init or an update script discover things like name, version, etc. | 12:27:43 |
| @blaggacao:matrix.org | The ecosystem stance 'mir san mir und die anderen san Frisöre' hasn't aged well. | 12:28:20 |
| @blaggacao:matrix.org | So ecosystem interoperability is in-scope for the Nix Ecosystem. | 12:29:29 |
| figsoda | really looking forward to pkgs modules and and simple package paths | 13:03:08 |
 @kranzes:matrix.org | I mean is pkg modules actually gonna be a thing for nixpkgs? | 13:04:19 |
| @kranzes:matrix.org | Is the plan to convert all nixpkgs to that? | 13:04:39 |
| @kranzes:matrix.org | That seems almost impossible | 13:04:45 |
| figsoda | there is a working group, i assume this is something they want to explore | 13:04:45 |
| @kranzes:matrix.org | Rfc140 though is nice though | 13:05:12 |
