Proposer Commitment / Preconf Policy

[DrewVanderWerff]

Thoughts and Principles on any Proposer Commitment / Preconf Policy

I am working on an open-source public good project called Commit-Boost (please see here for an introduction). We received a grant from Lido (thank you again for this), have had significant investment from NOs that are part of Lido (for testing, development, and research) and have gathered broader support from the Ethereum community in a variety of functions.

Over the last few weeks, I have seen posts and comments alluding to a new DAO policy surrounding proposer commitments and preconfs, and with this in mind, wanted to post some thoughts for principles that the Lido DAO may consider if they do create a policy. I hope readers find these principles helpful and support a path to (in no specific order) a thriving market of proposer commitments as well as keeping Etheruem’s validator set robust and decentralized.

A lot of this thinking is grounded in feedback from the Lido Community and others across Ethereum who have been instrumental in how Commit-Boost has been structured. Also, in my view, we are very early in all this, so would ask the Lido Community and DAO to caution against putting out a policy at this point and would opt in for giving NOs maximum flexibility until the market can develop.

1. Avoid governance where possible: Minimize governance as much as possible, and if required, focus governance in one part of the stack (i.e., on the types of commitments and what commitments validators can make versus how or who they make them through).

2. Caution exclusivity / restrictiveness: The policy should advocate for increasing decentralization which includes optionality. This is key for a market to flourish that not only puts Lido and NOs / Etheruem’s validators in the driver’s seat to let the market develop and service their needs, but encourages innovation and competition. If the policy excludes certain teams, clear measurable metrics on how they can get moved out of exclusion would be important (i.e., leave as little up to interpretation as possible and make sure these metrics are well discussed among NOs + community before implemented). With that said, I would try and avoid this as much as possible as it can cause unique market dynamics.

3. We don’t know everything today: The policy should be as future proof as possible. For instance, would this policy allow for new ideas to flourish such as out-of-protocol inclusion lists (i.e., IL-Boost).

4. Gather feedback from the community and make the discussion meetings + decisions fully transparent: If a policy is required, would highly encourage Lido to develop this policy in the open and allow for comments from the community, ideally with input from folks across Ethereum (including researchers, NOs, teams building modules, infra providers, etc). At a minimum, the goal should be to gather feedback from across a majority of NOs supporting Lido.

5. Use independent third parties to assess risk: This should be on an ongoing basis. Consider timely actions with decisions and as risk shifts. It is critical that this is relying on independent third parties such as Rated to help make quantitative assessments (similar to credit rating agencies).

6. Control for conflicts: If the Lido Community goes down the route of governance by Committee, would encourage ensuring that this Committee is composed of folks that represents the Ethereum Community as these decisions can have broad implications and folks who are conflicted to recuse themselves from decision making.

7. Trust the NO’s that makeup Lido: From all my interactions with Lido NOs, they are the some of the most thoughtful and robust operators in Ethereum.

I appreciate the consideration on these from the Lido Community and am excited by Lido’s willingness to lean in and support innovations such as preconfs.

[skozin]

Hi Drew, thanks for bringing up an important topic for discussion!

I’m pretty removed from social media lately so I can’t say anything about posts you refer to but let me share my personal opinion on some of the points you bring up in the proposal.

I fully agree that the DAO should minimize governance where possible, and all contributors I know are very supportive of it. However, currently, governance is the only gadget that aligns the interests of different groups of protocol users (stakers and NOs) as well as the interests of the wider Ethereum community. It’s not the only possible gadget, e.g. there are designs that allow stakers to more directly express their preferences by choosing a node operator or even a specific validation setup to stake with, but, as of now, the protocol relies on governance for achieving this alignment. How NOs validate the network significantly affects all of the above parties so I’d argue that ultimate governance minimization in this aspect of the protocol is only achievable by introducing these alternative gadgets (which btw I think is a good direction to consider and research).

One major objective of governance is risk minimization for both stakers and the network, and policies serve an important role here. So I’d be very cautious with the idea that NOs should be given maximum flexibility in running any validation setup without carefully considering 1) potential risks a particular validation setup brings, 2) whether the resources the DAO and its contributors have are adequate to roll out a particular change in policy at a given pace without significantly increasing the risks.

So, imo, there’s a balance between a policy being future-proof and safe which should be considered separately for each policy change. This is true for any policy, not only proposer commitments one.

That said, I’m totally with you regarding the need for and importance of transparency in policy-making, gathering feedback from the community (including, importantly, NOs), independent risk assessment, and control for potential conflicts.

And, last but not least, I think commit-boost is a great development that has the potential to diversify Ethereum validation setups and thus remove single points of failure.

Above is my personal opinion; I’m sure there will be divergent views within the community, and hope those interested in the topic will participate in the discussion.