Should the SC vote for confidence instead of for no-confidence?

[wolfgangwalther]

Question

The constitution reads:

A simple majority within the SC may call a reelection of the entire SC based on perceived loss of confidence.

This can lead to a 3-3 deadlock as we currently see. In this deadlock, neither the reelection, nor any majority vote can pass - the SC is unable to function.

In many legislatures, this is flipped into a vote of confidence that must pass to avoid a re-election. This would have the same results for a full SC, but avoid situations like the current 3-3 deadlock.

Do you support amending the constitution accordingly?

Candidates I'd like to get an answer from

@GaetanLepage
@djacu
@getchoo
@mschwaig
@pinpox
@tomberek

[nixos-discourse]

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

https://discourse.nixos.org/t/call-for-full-re-election-of-the-steering-committee/70208/7

[nix-elections]

@nixos-discourse: As you are currently not an eligible voter, please abstain from the election processes. You may consider requesting an exception.

[pluiedev]

Honestly I'm not sure why we didn't have a clause like that in the constitution to begin with. You'd think that if the SC doesn't have the confidence of >=50% of its members it would be acting without majority consent, right? Hindsight is perhaps 20/20 here...

It's also important the SC be capable of governing and managing its own affairs instead of delegating that task to other organs. If an SC that couldn't even manage to figure out how to manage itself is elected, we will all be in deep, deep trouble.

[philiptaron]

Yes, I support turning the mechanism into a vote of confidence through an amendment to the constitution.

Edit 2025/10/02: Having read @JulienMalka's finessed version, that's even better, and supports stability in a way the previous version did not.

[crertel]

I believe the correct thing, in the case of deadlock, would be to allow the Board to cast the deciding vote--it's a boring solution that requires minimal involvement beyond the initial amendment.

---

Vote Boring Nix.

[niklaskorz]

I support turning the mechanism into a vote of confidence through an amendment to the constitution.

[adamcstephens]

Yes, I support this change. I also think we need a requirement for replacing members, as having a full group is best.

[7c6f434c]

Procedural note: removed a reply-to-reply

[K900]

Agreed, though I'd also like to see a process for filling empty seats quickly, as there will likely be deadlocks that are not worth reaching for the measure of last resort.

[Scrumplex]

I support this proposal

[jopejoe1]

I'm in favor of turning it into a vote of confidence.

[nyabinary]

I 100% support this proposal; yes.

[drupol]

I am in full support of this proposal.

The current no-confidence mechanism, which requires a simple majority to trigger a new election, presents a clear and significant risk of deadlock. As pointed out, a split vote (e.g., 3-3) could paralyze the Steering Committee, leaving it unable to function or resolve its own internal gridlock, which would be detrimental to the entire NixOS project. Which is basically what happened.

Switching to a vote of confidence is an elegant and robust solution to this problem. It reframes the process from a reactive "vote to dissolve" to a proactive "vote to affirm" the committee's mandate. If a majority cannot actively express confidence in the current SC, it's a clear signal that a new election is necessary. This is a well-established and effective governance model used in many parliamentary systems for exactly this reason: it prevents stagnation.

It's encouraging to see the strong consensus in the comments. This change, would be a significant step forward for the stability and long-term health of NixOS governance.