qt6.qtdeclarative: re-enable qml caching

[outfoxxed]

Description of changes

Re-enables qml caching.

The QML tag ensures only the same version of qtdeclarative that generated a cache can load it, and the qmlc file names are now hashed with their store path to ensure they are unique. AOT compiled / bytecode caches have no versioning issues to worry about and are also re-enabled.

Potential issues:

• Currently we avoid qmlc collisions by including the store path of the application in the qmlc hash. This is problematic because it won't replace old caches when the derivation changes, just make new ones. I'm holding off on making changes to the structure of qmlc files until I get some input on this, but I think not cluttering the user's cache with outdated files is a good idea when possible.
• Replacing the qmlc version tag may not actually be required, especially because qt mirror source code should have a .tag file. We should drop that commit if possible.
• I'm not sure what the patch file convention is supposed to be, as there are a bunch of both numbered and non numbered patches in the qt patches folder.

I've tested this with a couple kde programs and made roughly sure it works. Further testing will be done for a final revision.

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/)
• 24.11 Release Notes (or backporting 23.11 and 24.05 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.

[outfoxxed]

Link back to ongoing matrix discussions https://matrix.to/#/!rGlCMuXgAhgEpdvJUz:nixos.org/$82acJFFrxAi8AguZYi2ZS74po0YMyI_w7SoyVenuxzU?via=nixos.org&via=matrix.org&via=kde.org

[outfoxxed]

@K900 Made the changes we discussed. The hashes are now included in the library hash field for nix store paths.

I've opted to just refuse to cache bare qml files on nix store paths for now, but embedded qml files and non store files will be cached.

[outfoxxed]

*Fixed error messages on mismatch

[outfoxxed]

*fixed conflicts with changes on master

Note: patch has been tested for: mismatched qtdeclarative, mismatched applications, qml on bare store paths.

[outfoxxed]

Should I add myself to qtdeclarative's maintainers for this patch as well?

[K900]

If you're willing to, sure.

[outfoxxed]

added, otherwise unchained

[nix-owners]

The PR's base branch is set to staging, but 372 commits from the master branch are included. Make sure you know the right base branch for your changes, then:

• If the changes should go to the master branch, change the base branch to master
• If the changes should go to the staging branch, rebase your PR onto the merge base with the staging branch:

  # git rebase --onto $(git merge-base upstream/staging HEAD) $(git merge-base upstream/master HEAD)
  git rebase --onto fa22dc9e8e3be834f7da92223a0f03767468e151 cd34221a4b06d3b9af727f18322a45af1c787733
  git push --force-with-lease

[SuperSandro2000]

@K900 we want to merge this?

[K900]

Next cycle.

[outfoxxed]

@K900 we want to merge this?

Note that I haven't tested the patch against 6.8.1 yet, though no changes were required.

I was thinking I would wait for staging-next to have more dependencies done before testing it, but just remembered I have an environment for 6.8.1 on top of unstable already. I can retest the patch itself there tomorrow, though I can't test the nix expression.