AIP: Nova Fee Router Proposal (ArbOS 30)

Constitutional Proposal

Abstract / Motivation

Currently, the DAO receives transaction fees on Nova at an address accessible via the core governance system. This means that any time the DAO wishes to transfer these funds, it must pass a constitution proposal. We propose instead a system in which these funds are “automatically” (permissionlessly) sent from Nova to the Arbitrum DAO Treasury on Arbitrum One. This gives the DAO faster, easier accessibility to these funds, and should make bookkeeping simpler.


Currently, the portion of the Nova transaction fees that the DAO collects are sent via the fund distributor contracts to 0xf7951D92B0C345144506576eC13Ecf5103aC905a, the address alias of the core governance l1 timelock. This allows the DAO to spend the funds via a round-trip core proposal initiated from the core governor on Arbitrum One.

A system in which these funds were sent to the DAO Treasury without the DAO needing to take action is preferable for several reasons:

  • Spending from the DAO treasury requires a lower quorum (3% vs. 5%)
  • Spending from the DAO treasury is faster (doesn’t impose the additional ~ 2 weeks of delay
  • Keeping all fees in a single location makes bookkeeping simpler and easier to reason about.

We propose an update in which all Nova fund distributors that distribute to the timelock alias are updated to instead distribute to a system of “fee routers”, via which permissionless transactions will allow the funds to be routed to the Arbitrum Treasury.

The lifecycle works as follows:

  1. A distributeRewards is called on a RewardDistributor contract, sending funds to a ChildToParentRouter contract.
  2. Either upon receiving funds or via a call to routeFunds, the ChildToParentRouter creates an L2-to-L1 message which sends the contract’s full Ether balance.
  3. The L2-to-L1 message is executed, transferring the Ether to a ParentToChildRouter contract on L1.
  4. routeFunds is called on ParentToChildRouter, creating a retryable ticket which transfers its full Ether balance to the DAO Treasury on Arbitrum One.


Additional Action

Sometime after this proposal action, an additional proposal can be submitted to transfer the funds remaining in the timelock’s address alias to the fee router.


If this vote passes on Snapshot, the tentative plan is to upgrade Arbitrum Sepolia to arbOS 30 the week of June 10th.

If Arbitrum Sepolia upgrades to arbOS 30, there will be an ecosystem-wide plan to ensure non-breaking compatibility of infrastructure and applications with Stylus, which will take several weeks or months. The ultimate goal is that the experience for EVM developers remains as unchanged as possible if arbOS reaches mainnet. Once enough confidence is reached, there will be an onchain vote to upgrade Arbitrum One and Nova to arbOS 30. The Arbitrum community will be informed of any significant updates as they arise.

1 Like

Thanks for flagging. I’m in favor of an automatic/permissionless system to make this transfer.

Hey everyone, a reminder that we’ll be hosting a short governance call later today (Monday, 29 April) at 14:30 UTC to walkthrough this proposal, and answer questions if any.

Link to join:

Hey everyone, here’s the recording of the governance call - Gov Call for RIP7212 and Nova Fee Router proposals (2024-04-29 22:35 GMT+8) - Google Drive

We are in favor of this and see no reason as to why this would be voted down. We wonder if there are any other situations like this where fees or revenues get sent to an address that is unaccessible outside of a governance vote? Could be worth changing those over as well.

This proposal gives the DAO faster, easier accessibility to these funds, and should make bookkeeping simpler. This has important implications for DAO governance. I’ve voted FOR!

The idea is good, I would like to eliminate unnecessary voting on obvious issues.

However, I would like to clarify whether someone will be checking for contract changes?

Voting “For”. This will reduce admirative burden of the DAO and I agree with the proposal that it will be faster, easier and simpler to handle the Nova fees in this manner. At no cost, I see no reason to vote against something like this.

No reason to vote against this. Having funds being sent in a permissionless way to the dao treasury instead of having to vote for it should be a trivial decision.
Voting for.

Voting “For” to turbocharge fee transfers to the Arbitrum DAO Treasury, streamline our finances, and slash the red tape of constant constitutional proposals!

1 Like

I voted FOR this proposal on Snapshot because it simplifies the operations of the DAO without any significant drawbacks.

I voted in favor. As it’s mentioned, the proposal will provide faster, easier and simpler access to these funds. I only see benefits from its implementation.

On behalf of the UADP:

We have voted Yes. There are basically no significant drawbacks and it helps simplify the DAO’s operations. The easier it is to return/send revenues to the treasury the better!

Blockworks Research will be voting FOR this proposal on Snapshot.

Supporting this change seems obvious to us as there are clear benefits and, to our understanding, no notable concerns or drawbacks.

Voted for. The automation of this process is a good practice and make the whole system more resilient.

Voted “For” to expedite funding when needed. No reason why this should not be automatic.

Voted “For”.
From Nova to the Arbitrum DAO Treasury, enhancing accessibility and efficiency. By automating these transfers, the DAO benefits from faster fund availability, reduced quorum requirements, and simplified bookkeeping, thereby improving overall governance and operational effectiveness.

Michigan Blockchain is voting for. This solves unnecessary voting and simplifies DAO operations.

We support the implementation of this proposal as it should greatly enhance the DAO’s operational efficiency.

Voting FOR this proposal, foundation made an incredible effort to ensure we all understand the importance of these updates to the protocol.