auto-rebase: add -f to the filter-branch invocation

[Ma27]

...because it tells me to do so:

+ git filter-branch --parent-filter 'sed '\''s/2140bf39e41767f25a395d20fb0d5698b8934b33/374e6bcc403e02a35e07b650463c01a52b13a7c8/'\''' --tree-filter 'nix-shell --run treefmt' 374e6bcc403e02a35e07b650463c01a52b13a7c8..HEAD Cannot create a new backup. A previous backup already exists in refs/original/ Force overwriting the backup with -f

If this is the wrong thing to do, feel free to propose a better solution. I just want this to work somewhat reliably.

Things done

• Built on platform:
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• Tested, as applicable:
  • NixOS tests in nixos/tests.
  • Package tests at passthru.tests.
  • Tests in lib/tests or pkgs/test for functions and "core" functionality.
• Ran nixpkgs-review on this PR. See nixpkgs-review usage.
• Tested basic functionality of all binary files, usually in ./result/bin/.
• Nixpkgs Release Notes
  • Package update: when the change is major or breaking.
• NixOS Release Notes
  • Module addition: when adding a new NixOS module.
  • Module update: when the change is significant.
• Fits CONTRIBUTING.md, pkgs/README.md, maintainers/README.md and other READMEs.

---

Add a 👍 reaction to pull requests you find important.

[Ma27]

That beign said, I get unrelated conflicts in Nix snippets in markdown files. So it seems the treefmt.toml is somewhat off?

[Ma27]

Yeah, the first commit from the patchset I rebased contains every single Nix-in-markdown snippet reformatted 🤡

[wolfgangwalther]

Yeah, the first commit from the patchset I rebased contains every single Nix-in-markdown snippet reformatted 🤡

This means you likely ran an outdated version of treefmt, not from the latest master branch. Formatting these codeblocks was introduced in #427460.

Edit: Although... if you get the reformatting for the codeblocks, but their format changes.. that looks more like "latest treefmt, but outdated nixfmt". You might need to run all of these inside a nix-shell, so that the nixfmt available is 1.0.0. It could be that you ran it outside a nix-shell and had system-wide nixfmt installed from nixfmt-classic or so?

[Ma27]

Edit: Although... if you get the reformatting for the codeblocks, but their format changes.. that looks more like "latest treefmt, but outdated nixfmt". You might need to run all of these inside a nix-shell, so that the nixfmt available is 1.0.0. It could be that you ran it outside a nix-shell and had system-wide nixfmt installed from nixfmt-classic or so?

Probably. I guess I should revise my entire dev setup for nixpkgs now to avoid further of these issues.

The change itself should still be valid, no? @wolfgangwalther

[wolfgangwalther]

tbh, I don't know whether that's the correct change or whether it should not even get there, so something else would be off. I have never used that script myself either.

[Ma27]

ping @infinisil then.

[Ma27]

ping @NixOS/nix-formatting

[dyegoaurelio]

I didn't manage to reproduce the issue

What's the diff you're actually encountering in the markdown files?
What branch are you trying to rebase (to and from) ?

[infinisil]

Iirc the script also failed for me when I actually wanted to use it in practice. Pretty disappointing because I even wrote a test for it :/

Regardless, this change is fine imo

[jfly]

Honestly have very little context on this codepath, but it seems low-risk to me, as we have branch protection enabled in this repo.

[MattSturgeon]

Personally I'd prefer to use verbose CLI flags in script sources, as they add additional self-documenting context.

Suggested change

	FILTER_BRANCH_SQUELCH_WARNING=1 git filter-branch -f \
	FILTER_BRANCH_SQUELCH_WARNING=1 git filter-branch --force \