inconsistent SRI hash requirements?

[vcunat]

Describe the bug

In SRI hashes, sometimes the final = is required and sometimes it is not.

Our nixpkgs have many cases where it's omitted, e.g.

pkgs/by-name/au/autorandr/package.nix:    hash = "sha256-8FMfy3GCN4z/TnfefU2DbKqV3W35I29/SuGGqeOrjNg";

because nix doesn't protest when building these.

But some commands do require it, e.g.

$ nix --extra-experimental-features nix-command \
  hash convert --from sri sha256-8FMfy3GCN4z/TnfefU2DbKqV3W35I29/SuGGqeOrjNg
error: input hash 'sha256-8FMfy3GCN4z/TnfefU2DbKqV3W35I29/SuGGqeOrjNg' does not have the expected format '--from sri'

Expected behavior

Right now we can't use the command (easily) to process nixpkgs, etc. So probably allow missing = in the command.

• checked latest Nix manual (source)
• checked open bug issues and pull requests for possible duplicates

---

Add 👍 to issues you find important.

[vcunat]

As a workaround, one can use e.g. nix-hash, though that produces an incorrect warning :-(

[vcunat]

Oh, dropping --from sri is a workaround that works. Really confusing to me that it makes a difference, as it obviously can't match any other format but SRI.

[nixos-discourse]

This issue has been mentioned on NixOS Discourse. There might be relevant details there:

https://discourse.nixos.org/t/2024-12-04-nix-team-meeting-minutes-200/57005/1