bochs: move to by-name hierarchy

[AndersonTorres]

Description of changes

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/)
• 23.11 Release Notes (or backporting 23.05 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.

[infinisil]

Oh this is actually a bit concerning. This is a breaking change because .override { wxGTX = ...; } doesn't work anymore. I don't think we should make these changes until we have a better story for that, even if this allows removing definitions from all-packages.nix.

Note that even if you need custom arguments, you can still use the pkgs/by-name directory.

[AndersonTorres]

Bochs didn't update their dependencies yet (they are more concerned about optimizing the emulation core), so overriding wxGTK to other version than that specific one will not work.
I considered the "all-packages-mode" but it didn't feel useful.

[infinisil]

Overrides can also be used to just make minor changes to packages, e.g.

.override (old: {
  wxGTK = old.wxGTK.overrideAttrs ...;
})

Similarly the darwin change could also cause breakage, consider:

.override {
  libobjc = ...;
}

These overrides would now silently not do anything anymore.

Of course, Nixpkgs has always not really enforced not having such breakages, but the migration to pkgs/by-name should not cause more of these to happen. This is a problem to be fixed in the future.

I opened #254632 to have this written down.

[AndersonTorres]

Similarly the darwin change could also cause breakage

That Darwin-specific things were always problematic to me, since the splicing incident (that, among other things, break inherit (X) a b c; design pattern).

[infinisil]

Oh yeah that also messes with it, there's definitely many things to be improved. I feel like the custom argument one might be the next one to dig into.