woodpecker-{agent,cli,server}: init at 0.15.3

[ambroisie]

Description of changes

Supersedes #169470.
Closes #178332.

Some notes :

• We probably want to prefix the different binaries with woodpecker-, instead of just cli, agent, and server.
• Address/remove the FIXME about musl.
• Perhaps add an update script.

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandbox = true set in nix.conf? (See Nix manual)
• Tested, as applicable:
  • NixOS test(s) (look inside nixos/tests)
  • and/or package tests
  • or, for functions and "core" functionality, tests in lib/tests or pkgs/test
  • made sure NixOS tests are linked to the relevant packages
• Tested compilation of all packages that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage
• Tested basic functionality of all binary files (usually in ./result/bin/)
• 22.11 Release Notes (or backporting 22.05 Release notes)
  • (Package updates) Added a release notes entry if the change is major or breaking
  • (Module updates) Added a release notes entry if the change is significant
  • (Module addition) Added a release notes entry if adding a new NixOS module
  • (Release notes changes) Ran nixos/doc/manual/md-to-db.sh to update generated release notes
• Fits CONTRIBUTING.md.

[wyndon]

Do you plan to add a module?

[ambroisie]

Do you plan to add a module?

I plan on migrating my own configuration from Drone to Woodpecker, if nothing breaks in that process.

I don't think I'll be adding a module (in this MR at least), mostly because one might want to have multiple instances of the agent running at once on the same machine (which is not supported very well by the module system), e.g: a docker backend and an exec backend.

[ambroisie]

Not sure why ofborg is unhappy about (what seems to be) the mkYarnPackage derivation.

[06kellyjac]

it's because yarn2nix is reading the package json into eval, so it's a build output -> eval aka IFD

https://github.com/NixOS/nixpkgs/blob/master/pkgs/development/tools/yarn2nix-moretea/yarn2nix/default.nix#L176

you'll want to copy the package json over into the repo. or I can see about removing the IFD issues

[ambroisie]

@06kellyjac I would rather not copy the package.json file since upstream is doing everything correctly, and it would just be one more hurdle when updating the package.

My wondering is more along the lines of why other packages that make use of mkYarnPackage either don't have that issue with ofborg, or if they do then why hasn't it been fixed/changed?

[06kellyjac]

they all just copy in package.json, its a simple wget or curl in the update script. Again I'll see if I can tweak yarn2nix to stop the IFD issues when I get some time

---

I made some changes e243a8c

• adds prefixes to bins
• removes the need for libexec/xyz etc by just keeping the dist folder
  • before: λ du -h -d 0 ./result/ -> 359M
  • after: λ du -h -d 0 ./result/ -> 664K
• seperate the frontend to its own input, this is important so people can fix it if it breaks like jellyfin and all other node stuff did recently jellyfin-web does not build on master #176288
• drop the extra static stuff, it seems to cause more issues than it solves, we don't need to exactly copy the upstream builds
  • LMK if this resolves the glibc musl stuff, not sure what the actual issue was in order to test it

server started and ui shows fine but that's all I did:

WOODPECKER_GITHUB=true ./result/bin/woodpecker-server --server-host http://192.168.0.111

[ambroisie]

@06kellyjac I think we should only keep the front-end as a passthru attribute, I don't think it makes much sense to have it as an actual package in nixpkgs but it would still allow overriding thusly (approximately? not tested):

self: super:
{
  woodpecker-server = super.woodpecker-server.override {
    woodpecker-frontend = super.woodpecker-server.passthru.woodpecker-frontend.override {
      # Override here
    };
  };
}

[06kellyjac]

I'm not super familiar with passthru but yeah since it's an input to server you could replace it with anything, in this case an overriden frontend grabbed from passthru

so yeah that looks fine

---

As for the remaining eval/IFD issues I think it's because you need to set offlineCache or it just does it magically with IFD stuff

[ambroisie]

Pushed some fixups (which will need to be rebased/squashed).

[wyndon]

Result of nixpkgs-review pr 178441 run on x86_64-linux 1

3 packages built:

• woodpecker-agent
• woodpecker-cli
• woodpecker-server

[nixos-discourse]

This pull request has been mentioned on NixOS Discourse. There might be relevant details there:

https://discourse.nixos.org/t/prs-already-reviewed/2617/531