flashbench: 2012-06-06 -> 2020-01-23

[Luflosi]

Motivation for this change

The current version of flashbench is quite old. I updated it to use the latest commit from about a year ago. There have only been four new commits with small changes since 2012, so I don't expect any regressions.
I also switched from fetchgit to fetchGitHub.

Things done

• Tested using sandboxing (nix.useSandbox on NixOS, or option sandbox in nix.conf on non-NixOS linux)
• Built on platform(s)
  • NixOS
  • macOS
  • other Linux distributions
• Tested via one or more NixOS test(s) if existing and applicable for the change (look inside nixos/tests)
• Tested compilation of all pkgs that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review wip"
• Tested execution of all binary files (usually in ./result/bin/)
• Determined the impact on package closure size (by running nix path-info -S before and after)
• Ensured that relevant documentation is up to date
• Fits CONTRIBUTING.md.

[r-rmcgibbo]

Result of nixpkgs-review pr 112897 at ee7a1d1b run on x86_64-linux 1

1 package built:

• flashbench

2 suggestions:

• warning: missing-phase-hooks

  installPhase should probably contain runHook preInstall and runHook postInstall.

  Near pkgs/os-specific/linux/flashbench/default.nix:14:3:

     |
  14 |   installPhase = ''
     |   ^
  

• warning: unclear-gpl

  gpl2 is a deprecated license, check if project uses gpl2Plus or gpl2Only and change meta.license accordingly.

  Near pkgs/os-specific/linux/flashbench/default.nix:25:5:

     |
  25 |     license = licenses.gpl2;
     |     ^
  

[r-rmcgibbo]

Result of nixpkgs-review pr 112897 at ee7a1d1b run on aarch64-linux 1

1 package built:

• flashbench

2 suggestions:

• warning: missing-phase-hooks

  installPhase should probably contain runHook preInstall and runHook postInstall.

  Near pkgs/os-specific/linux/flashbench/default.nix:14:3:

     |
  14 |   installPhase = ''
     |   ^
  

• warning: unclear-gpl

  gpl2 is a deprecated license, check if project uses gpl2Plus or gpl2Only and change meta.license accordingly.

  Near pkgs/os-specific/linux/flashbench/default.nix:25:5:

     |
  25 |     license = licenses.gpl2;
     |     ^
  

[Luflosi]

I changed the license from gpl2 to gpl2Only.

[Luflosi]

Should I add the runHooks to the installPhase? I'm not convinced, it's actually useful here.

[SuperSandro2000]

Should I add the runHooks to the installPhase? I'm not convinced, it's actually useful here.

Otherwise the phases are not run in overlays which is usually expected.

[Luflosi]

Thanks for the explanation, I added them. Please review.

[rycee]

Thanks! Squashed and rebased to master in ba68041. Note, I moved the "unstable" to pname since it is part of the name, not the version.

[SuperSandro2000]

Thanks! Squashed and rebased to master in ba68041. Note, I moved the "unstable" to pname since it is part of the name, not the version.

Thats wrong. We only have one flashbench package so this part belongs to the version.

[rycee]

I disagree, it is always part of the name:

nix-repl> builtins.parseDrvName "flashbench-unstable-2020-01-23"
{ name = "flashbench-unstable"; version = "2020-01-23"; }

It may make intuitive sense to put the "unstable" in the version field but I think this is misleading since it doesn't match the package name and version that Nix sees.

[SuperSandro2000]

It may make intuitive sense to put the "unstable" in the version field but I think this is misleading since it doesn't match the package name and version that Nix sees.

That function is by design broken and needs a fix. The last hundred or so PRs I reviewed where done with the unstable in the version number because people tend to use pname all over the place.

See #68518 and https://repology.org/projects/?search=flashbench which follow no version number at all.

[rycee]

I agree that the unstable naming is broken but I believe fixing it by diverging from the Nix semantics w.r.t derivation names is bad. I've added a comment elsewhere (#110442 (comment)) to this effect and propose to switch the discussion there, thank you for pointing out the other issues.