copilot-language-server: add fhs version

[arunoruto]

Solves issue #391730 and closes the PR #391748.

The main inspiration for the FHS environment were the vscode and zed-editor packages. An fhs passthru has been created and later referenced in pkgs/top-level/all-packages.nix.
I am not quite sure, if this is the correct of doing this, but it was the only reference I had.

This PR will provide an additional package copilot-language-server-fhs, which should solve the issue addressed in #391730. The "normal" package is kept, since it works on stable releases so far. It is possible, when 25.05 is released, that the normal package will break. The FHS package will then become the default one in the future.

I would be happy if someone could provide a better solution or a fix to the underlying problem. I guess it isn't packages correctly or the package needs additional information?

As for the PR #391748: using the auto patch hook breaks the binary and AFAIK it can't be fixed afterwards, which makes FHS environments the only solution.

If someone has the same problem as addressed in the issue, please test this PR and see if it solves your problem. Looking forward for your feedback.

For those who have a merge/commit bit: please only merge this PR after some testing has been done!

TODO:

• It would be nice to have a versionCheckHook too, but I am not sure if I should introduce this in this PR or make a separate one.

Ping: @wattmto @drupol @brianmcgillion

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandboxing enabled in nix.conf? (See Nix manual)
  • sandbox = relaxed
  • sandbox = true
• 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/)
• 25.05 Release Notes (or backporting 24.11 and 25.05 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
• Fits CONTRIBUTING.md.

---

Add a 👍 reaction to pull requests you find important.

[arunoruto]

Running both commands on my system via my own flake, not this PR:

../flake ❄ ❯ nix run .#copilot-language-server --impure -- --version
1.290.0

../flake ❄ ❯ nix run .#copilot-language-server-fhs --impure -- --version
1.290.0

nixpkgs-review result

Generated using nixpkgs-review.

Command: nixpkgs-review pr 393260

---

x86_64-linux

✅ 1 package built:

• copilot-language-server-fhs

[drupol]

You can add the versionCheckHook within this PR, in a new commit.

[brianmcgillion]

@arunoruto thank you for all your efforts on this, really appreciated. can confirm that the -fhs version is working.

[arunoruto]

You can add the versionCheckHook within this PR, in a new commit.

I add the versionCheckHook in the main package, not sure if it can be also added in the FHS env.

[drupol]

You can add the versionCheckHook within this PR, in a new commit.

I add the versionCheckHook in the main package, not sure if it can be also added in the FHS env.

Have you tested it before pushing? Do you see the output in the build log that it has been working successfully ?

[arunoruto]

Minor changes:

• Included a version attribute in buildFHSEnv
• Replaced copilot-language-server argument name with package
• Replaced the reference to the variableexecutableName with package.meta.mainProgram, so the fhs function only depends on the argument package.

I will probably make a new PR and try to include testers.testVersion instead of versionCheckHook, with the hope to catch weird behaviour with the binary as introduced by #391730:

Note

This is a check you add to passthru.tests which is mainly run by OfBorg, but not in Hydra. If you want a version check failure to block the build altogether, then versionCheckHook is the tool you’re looking for (and recommended for quick builds). The motivation for adding either of these checks would be

• Catch dynamic linking errors and such and missing environment variables that should be added by wrapping.
• Probable protection against accidentally building the wrong version, for example when using an “old” hash in a fixed-output derivation.

And thank you @drupol for reviewing the PR and for the tips!

nixpkgs-review result

Generated using nixpkgs-review.

Command: nixpkgs-review pr 393260

---

x86_64-linux

✅ 2 packages built:

• copilot-language-server
• copilot-language-server-fhs

[benley]

(rebased to pick up the latest version of the package from nixpkgs)

This seems to work. Is there any reason we shouldn't merge it?

[arunoruto]

(rebased to pick up the latest version of the package from nixpkgs)

This seems to work. Is there any reason we shouldn't merge it?

Not that I am aware of. The passthru tests will be included in a separate PR since they would go beyond the topic of this one.

[drupol]

Hello,

I haven't merged this PR because I never packaged stuff related to fhs, this is why I'm waiting for someone else to review this PR.

Ping me in a couple of days, if nobody reviewed it, I'll merge it.

[SuperSandro2000]

Shouldn't we mention here what kind of trouble like that executables cannot be executed?

[arunoruto]

Could I reference the issue #391730 in the text? And if so, as link or just the issue number?

[SuperSandro2000]

A comment in the form of If you receive an error like: Could not start dynamically linked executable plus a link to that issue would be super cool :)

[arunoruto]

I added a disclaimer to open the package details for more information on when to use it, and added the error messages into longDescription. I wasn't sure if description supports markdown and it should be kept short, so I opted for the longDescription (hence the multiple force-pushes... check trice and push once 😂... still learning tho!).