The Future of Optimism Governance

The Optimism Foundation recently published this as a post on our Mirror blog. We’re reposting here in full for feedback and discussion from the community.

The Future of Optimism Governance

The Optimism Collective is governed by a two-house system. The Token House is made up of OP holders and their delegates, and the Citizens’ House is a one-person-one-vote system based on reputation.

This two-house design is intended to help the Collective make good decisions and avoid common pitfalls in token-based governance systems. Tokenholders represent one constituency out of many, and are not always the right unilateral owners for decisions that don’t involve the economics of the system. Creating two distinct branches of governance with different membership criteria allows the Collective to match constituents’ incentives with different types of decisions. It also helps the Collective avoid concentration of power, enact safeguards via checks and balances, and evaluate decisions from broader perspectives.

This is not a new concept. Bi- or tri-cameral governance systems have existed for centuries. With a new iteration of a known pattern, Optimism aims to build a governance system that will stand the test of time and help the Collective flourish.

This post provides an overview of the goals, design principles, and problem spaces of Optimism Governance. It is not a commitment or blueprint for exactly how governance processes will be introduced and implemented: it’s highly likely the Collective will learn from initial experiments and make modifications to the rules and processes outlined here. These “Problem Spaces” are intended as starting points for discussion.

Governance Goals

There are two primary goals of Optimism’s governance system:

  1. Capture resistance. Governance plays a key role in securing the anti-capture and censorship resistance of the Optimism protocol. Governance should (a) make it possible for chain or network operation to continue without reliance on any individual entity, and (b) prevent any one entity or small group of entities from being able to control or censor the protocol or its functions.
  2. Resource allocation. Governance’s second primary responsibility is to allocate resources effectively to support the Collective’s vision and accrue sustainable value to the Optimism Collective. Vision & value may often be in conflict, and allocating resources effectively involves a blend of short- and long-term thinking. This includes allocation of both the token treasury and protocol revenue.

Design Principles

Design decisions for the Collective’s governance system should be made in line with three key principles:

Governance minimization. The set of governance responsibilities that are encoded onchain or formalized in voting processes should remain as minimal as possible. The Collective aims to reduce governance to its essence and to avoid introducing regulation where freedom can achieve the same result. This principle is key to encouraging permissionless innovation. In practice, this looks like a minimal set of (1) onchain governance processes to upgrade Optimism contracts and tune the economic parameters of the system, and (2) offchain social processes to maintain a healthy community.

Iteration. Optimism is decentralizing iteratively to increase the chances of building a healthy system that lasts for the long-term. This means the Foundation will play a role in establishing processes, help the Collective through its first few rapid feedback loops in improving those processes, then reduce its role over time. (This also means the design principles and goals outlined in this document may be invalidated or updated along the way.) This iteration gives the Collective a chance to learn how to make thoughtful decisions using an un-intuitive but essential loop: introduce a governance process that involves active participation, then gradually work to automate or minimize it over time. Governance’s responsibility then becomes to adjust the autopilot when necessary, not to keep two hands on the wheel.

Forking. The right to fork and the right to exit are critical to protect individual freedoms. All of the core software and tooling required to run the Optimism network should be made open source, freely available, and easy to use such that a fork is always a viable alternative. This isn’t just about vibes: in crypto, where credible commitments not to extract are what makes decentralized platforms valuable, this is a competitive advantage. Participants will be more likely to join Optimism if they have the ability to make an alternative.

Balance. Influence in governance must extend beyond financial stake to value humanhood and intelligent life. The centralizing force of plutocratic token governance must be balanced with Citizenship. Giving voice and power to different constituents allows the Collective to match decisions with voters’ incentives, rather than building a solely plutocratic system. Checks and balances prevent capture.

Impact = profit. A key part of the Optimistic Vision is to ensure that every individual is rewarded in proportion to their positive impact to the Collective. We believe this to be the most important target in the pursuit of solving global coordination problems and creating a better future. Our economy is an expression of our collective needs, wants, and ethics. Optimism Governance is ultimately responsible for enacting this fairness ratio that is key to the Optimistic Vision.

Design Overview

The responsibilities of Optimism Governance can be categorized into one of the two primary goals of the governance system: Capture Resistance and Resource Allocation. Within each category, governance responsibilities are identified as (a) the exclusive responsibility of one or the other house, (b) a responsibility where one house drives proposals and the other house may veto, or (c) as a truly shared responsibility subject to approval by each house. This system is designed to help each house play to its strengths and avoid deadlock.

The Optimism Token House is made up of OP tokenholders and their delegates. Tokenholders and their delegates are expected to be rational economic actors interested in preserving or increasing the economic well-being of the Optimism Collective. Therefore, the Token House is responsible for decisions that affect business parameters of the system such as inflation or the variables that govern sequencer selection. Token voting is distributed widely and permissionlessly and helps represent the free market in the Collective’s decision making.

In contrast to the Token House, the Citizens’ House uses a one-person, one-vote system. Citizens are meant to represent individual human stakeholders of the Collective: builders, users, and community members who are aligned with the project’s values and are interested in the long-term benefit of the Collective. Therefore, the Citizens’ House is responsible for decisions that may require prioritizing long term growth ahead of other goals – decisions like allocating funding to public goods that support the Collective. In the long term, we can expect the Citizens’ House to expand to include the thousands or millions or billions of humans interacting with Optimism.

Governance responsibilities given to each house will change over time as the Collective iterates towards its long-term governance-minimized end state. This includes introducing highly manual governance responsibilities (e.g. inflation adjustments, or Superchain allowlists) that are intended to be replaced by automated smart defaults as the Collective learns how to govern these vectors appropriately. It may also include transitioning powers from the responsibility of the full governance body to a subset of governance participants who are empowered by their broader governance base.

In these early chapters of Collective governance, many of these iterations have been designed or implemented by the Optimism Foundation. Eventually this must transition to the responsibility of Collective governance itself. This meta-governance power marks the end of the Foundation’s stewardship of the evolution of Collective governance and entrusts the governance community with continued experimentation towards a healthy end-state for Optimism.

So: what does governance… well, govern?

Problem Space #1: Capture Resistance

Optimism Governance’s first responsibility is to ensure that no single party can unilaterally control, censor, halt, or otherwise extract rent from Optimism. A decentralized compute layer is hardly valuable if controlled by a centralized entity. Governance helps the Collective make decisions in the absence of any one person or organization that is “in charge.”

The responsibility to uphold capture resistance can be broken down into two broad categories:

  1. hard powers: governance powers that are eventually encoded and executed onchain, and
  2. soft powers: social norms the governance community upholds, not enforced onchain but by the community’s ability to fork.

1/ Hard powers

Protocol Upgrades refer to the ability for governance to make changes to the OP Stack protocol itself. This is a crucial responsibility of governance that ultimately makes the community the steward of Optimism’s core protocol functionality.

Protocol upgrades will be approved by the Token House, subject to a Citizens’ House veto. The Token House is the primary approver because of its broad reach and likelihood of representing business interests that depend on the OP Stack’s functionality. The Citizens’ House veto power guards against capture of the Token House and increases the resilience of the overall system.

Today, protocol upgrades are approved by the Token House and executed manually by the Optimism Foundation. In the coming Seasons, the Collective may implement a Security Council responsible for executing upgrades following the will of governance. The Citizens’ House veto power comes online in Season 5 (H1 2023), when the Citizens’ House comes fully online. Eventually, execution of protocol upgrades will be fully onchain. On a far enough time horizon, we can expect protocol upgrades to become less and less common as the protocol nears feature completeness. We can look to protocols like Ethereum and Bitcoin as a model for what mature decentralized protocol upgrades look like.

Sequencer Selection is the ability to make choices about which parties may act as sequencers for OP Chains or in shared sequencing schemes.

This begins as an explicit governance power driven by the Token House, with Citizens’ House power to veto. Sequencers are important economic actors for OP Chains and eventually for the Superchain. The Token House is well-incentivized to make sure sequencer selection is fair and beneficial for the Collective. The Citizens’ House veto power guards against the capture of the Token House related to sequencer selection.

Today, OP Chains are sequenced by a handful of different entities. In Season 5, those sequencers will submit proposals to Optimism Governance to be formally included in the Superchain, officially bringing the hard power of sequencer selection into the hands of the Collective. More information about this governance process will be shared in the coming weeks.

Eventually, sequencer selection may be fully automated based on verifiable and pre-agreed upon processes, enabling the sequencer set to be curated and maintained without active governance intervention.

Citizenship Eligibility is the responsibility of governance to select new humans to participate in the Citizens’ House of Optimism Governance.

Citizenship Eligibility will be determined by the Citizens’ House; the Token House will have the power to veto. Assuming good long-term values alignment with the Collective, Citizens should be incentivized to grow a set of voters that make quality decisions about RetroPGF allocations. A Token House veto guards against capture of the Citizens’ House or moves to consolidate power.

To date, Citizenship expansion has been administered by the Foundation. In the future, the Foundation will propose an expansion algorithm that Citizens vote to approve. Ultimately, Citizens may govern a citizenship selection algorithm directly that runs across identity data in the Optimism AttestationStation. In its final minimized form, the Citizens’ House has established a battle-tested onchain selection process; updates to the contracts that govern selection can be made as necessary by the Citizens’ House.

2/ Soft powers

Code of Conduct Enforcement is the responsibility of governance to uphold and enforce the Optimism Collective Code of Conduct.

Each house may govern the code of conduct for its own members. The best cultural and values-based enforcement for governance communities should come from within that specific community.

Today, CoC violations are processed with administration from the Foundation. In future Seasons, the community will iterate on this approach: Season 4 includes an RFP for Code of Conduct research, and Season 6 may introduce a new Council to replace the Foundation’s role in this process. As a soft power, Code of Conduct enforcement will not be implemented onchain. For example, if a Token House delegate were to violate the Code of Conduct, that delegate would not be prevented from voting by an update to Optimism’s voting contracts onchain. Rather, the community could enforce its rules and norms on the social layer by choosing to remove the participant from delegation frontends and UIs, or by informing the tokenholders who’ve delegated to the offending party of the breach in conduct.

Director Removal is the right of governance to remove a director of the Optimism Foundation. Director Removals are approved by the Token House.

This right is enforced and upheld by the Founding Documents of the Optimism Collective. As such, the Token House may vote to veto any reduction of this power. As Directors are not represented onchain, this responsibility does not have onchain enforceability.

Problem Space #2: Resource Allocation

The second primary responsibility of Optimism governance is to allocate the Collective’s resources to help it achieve its goals. This is a piece of essential governance that requires human intervention, especially in its early stages. Resource allocation in the Collective comes in three forms:

1/ Allocation of Protocol Revenue
2/ OP Treasury Management
3/ RetroPGF Funding

Allocation of Protocol Revenue is the responsibility of governance to determine how to direct surplus ETH generated by the Optimism protocol. Optimism generates revenue through transaction fees paid on OP Mainnet and other OP Chains. Part of these transaction fees is used to post data to Ethereum L1 and pay for other expenses associated with running the protocol, and the remainder of the fee is accumulated as surplus revenue that can be directed by the Collective for the benefit of the Collective. In the future, the sequencing and proving networks that help run the Optimism protocol may also provide sources of revenue for the Collective.

By default, surplus protocol revenue will be allocated to RetroPGF. The Token House may vote on proposals to divert a portion of the surplus protocol revenue for other Optimism Collective purposes as they see fit. These proposals will be subject to veto right by the Citizens’ House.

Protocol Revenue is a crucial lever in the Collective. In its early stages, the default allocation to RetroPGF lets the Collective reward projects that grow the Optimism ecosystem. As Optimism matures, the Token House is entrusted to propose deviations from this default for the health of the Optimism Collective. Citizens’ House veto powers help hold the Token House accountable and prevent short term optimization at the expense of long term sustainability.

In a future governance season, a proposal type will be introduced for proposing modifications to the allocation of surplus protocol revenue. These proposals may begin with certain safeguards, like a cap on possible allocations. Proposal power for high-level budgeting decisions may additionally be delegated to elected representatives as the Token House sees fit. The long term minimized state of this governance responsibility could include some thoughtful automation to determine revenue allocations based on indicators of whether the Collective should be in a “growth” stage or “value” stage.

OP Treasury Allocation is the responsibility of governance to direct how the existing OP token treasury is allocated. This has several components:

  1. The Governance Fund, which is directed by Token House governance.
  2. The RetroPGF Fund, which is directed by Citizens’ House governance, subject to a Token House veto.
  3. The Seed Fund and Unallocated portions of the token treasury, which are initially administered by the Foundation, and will eventually be directed jointly by both houses.
  4. Foundation Budget Approvals, which are governed jointly by both houses.
  5. Inflation adjustments, which are governed by the Token House, subject to a Citizens’ House veto.

This distributed ownership of the OP Treasury allows for flexibility and gives each house a clear mechanism to allocate tokens for the development of the Collective. Today, treasury allocations are administered by the Foundation, and will gradually move onchain as the governance system matures and becomes more resilient. In the long term, as most of the Treasury is circulating, this power will boil down to the governance of inflation, which can be automated to a thoughtful set of defaults based on the economic health of the Collective.

RetroPGF Grants are the responsibility of governance to allocate RetroPGF to projects that have provided public good to the Optimism Collective. These grants may refer to either (a) OP tokens from the RetroPGF Fund portion of the token treasury (see OP Treasury Allocation), or (b) Profits generated by the Optimism Protocol that governance has decided to allocate to RetroPGF (see Allocation of Protocol Revenue). Specifically, this responsibility includes:

  1. RetroPGF Scope: the ability of governance to update the scope, metrics, and impact evaluation criteria used in determining how RetroPGF Grants are allocated to projects. In line with the principle of governance minimization, this should be a soft power.
  2. Project Allocations: the ability of the Citizens’ House to make specific individual project-level decisions about how the RetroPGF Scope is applied to allocate grants. This is the corresponding hard power for RetroPGF.

Both components of RetroPGF Grants above are the responsibility of the Citizens’ House. Citizens are selected as individual humans who are interested in the long term health of the Optimism Collective, and RetroPGF is a long term investment in Optimism’s growth.

Today, the Foundation identifies RetroPGF scope and Citizens make Project Allocation decisions. Eventually, Citizens will govern RetroPGF scope themselves.


Altogether, this looks like the diagram below:

Additional Notes

Incompleteness

The list above does not include all governance powers – it notably excludes governance responsibilities, especially social ones.

The list above is also not a commitment or blueprint for exactly how governance powers will be rolled out and implemented. It is highly likely that the Collective will learn from initial experiments towards this blueprint and may make modifications to the processes outlined here. These “Problem Spaces” are intended as starting points for discussion, not as a prescriptive roadmap.

Checks, balances, and vetos

This post also excludes crucial design details about veto power, approval and quorum thresholds, proposal defaults, and failure modes. In particular, it’s important to note that all veto powers are not created equal. A very high veto threshold and quorum create a higher bar for exercising veto power, where a simple majority could make for a relatively easy veto. These thresholds determine the balance of power in responsibilities shared between houses and must be considered carefully.

The Collective will apply the principle of iteration in introducing these checks, balances, and vetos. This allows time to experiment with the right settings and make sure both branches of governance are developing safely.

These details are incredibly important to the design of a well-functioning system, and members of the Collective are consulting with a set of experts, from crypto and beyond, to design a quality set of patterns to increase the chances of Optimism Governance’s success. If you’re interested in contributing to these efforts, you’re welcome to reach out to bobby@optimism.io or chime in on the governance forums.

Citizenship and capture

Since Citizens play a crucial role in the Optimism governance, it’s paramount that the Citizens’ House is qualified and capture-resistant. Several factors come into play here, including the ways identity and contribution history can help inform citizenship eligibility; public and private voting strategies to help prevent collusion and bribery; citizenship incentives; escape hatches and metagovernance in the event of capture; and the iterative path towards growing the set of Citizens and their responsibilities. These important topics will be discussed in greater detail in a future post.

Key Milestones

Optimism Governance continues to make steady strides towards a robust and sustainable system. Here are some of the next milestones for the Collective in the coming quarters:

Season 5 [2024]

  1. Sequencer Allowlist / Law of Chains
    Season 5 will introduce the governance system’s power to welcome new Sequencers into the Superchain. More information about sequencer governance (and its relation to chain, user, and platform governance) will be shared in the coming weeks.

  2. Joint-House voting
    In Season 5, the Collective will bring the Citizens’ House fully online and execute its first set of joint-house votes. After executing three rounds of RetroPGF, this milestone will mark the next stage of maturity for the Citizens’ House as they participate in Optimism Governance beyond RetroPGF grant allocations.

  3. Security Council v1
    Season 5 will also include the Optimism Security Council. The Security Council is a set of community members tasked with executing protocol upgrades at the will of governance. This replaces the multisig wallet controlled by the Optimism Foundation and marks an important milestone in protocol decentralization.

Season 6 [2024]

  1. Protocol Revenue Allocation
    Season 6 will introduce the initial process for votes to allocate Protocol Revenue. It’s possible this process will include the introduction of a Treasury Council responsible for making high-level proposals to governance about how to allocate resources; any proposal for protocol revenue allocation will be submitted to the Token House for approval, and to the Citizens’ House for veto.

  2. Onchain treasury execution
    In Season 6, parts of the Optimism Treasury will move onchain. This means governance will be able to initiate transactions that move tokens from the OP Token Treasury entirely onchain. As noted above, this onchain execution will begin with an emergency safeguard that will be removed in future seasons.

30 Likes

A lot to take to the next citizens community call. I love how both houses are starting to work together on S5.

Sequencer decentralization looks like the bottle neck avoiding big capital to move from mainet to OP. Nice to see this problem being tackled.

I wonder if treasury and security councils will need any tool that coul be pushed with RFGs. @jackanorak ?

6 Likes

yeah, certainly there is stuff around facilitating intake and good decision making. in fact this should probably be handled directlyt hrough rfgs once we develop a good set of specs. to date a lot of analytics stuff has been a lot of “allocate us up front before there’s spec”

1 Like

this article is really helpful for a newbie like me in Optimism governance as it comprises a lot of information in one place.

looking forward to joint house voting and further technical decentralization.

3 Likes

Thank you for sharing @bobby

I’m happy that this post clearly outlines responsibilities of both houses and identifies clear scope for governance. I’m excited for what Season 5 and 6 have to bring :smiley:

  • Does the scope of DAO subcommittees fall under resource allocation?
  • Many have expressed interest in contributing to the RPGF scope, metrics or evaluations discussion:

Is there room in the RPGF governance to receive input from the token house and modify its parameters?

2 Likes

I am curious about transparency for Citizen’s house. Is there a public list of past/existing citizens and details on how they were selected?

1 Like

eth cc talk summarizing this overview plus some additional context: Justine Humenansky - Bootstrapping a bi-cameral governance system - YouTube

9 Likes

Thank you for the overview, so many exciting updates! I have a couple questions:

Can you give me an example of when this would be applicable?

What will be the selection process and criteria for the Security Council?

5 Likes

Does the scope of DAO subcommittees fall under resource allocation?

The Collective doesn’t really have subcommittees, but the existing Council model is a good example of how Token House can delegate decisionmaking authority (in the Grants’ Council’s case, over a portion of GovFund spend) to a subset of delegates.

Is there room in the RPGF governance to receive input from the token house and modify its parameters?

Certainly! For now the Foundation is learning from community feedback in iterating on each RetroPGF round. When this responsibility transitions to the Citizens’ House, we expect them to take input from the broader community as well.

2 Likes

I am curious about transparency for Citizen’s house. Is there a public list of past/existing citizens and details on how they were selected?

Citizenship for RetroPGF Round 2 was distributed using the process described here. More general info on Citizenship can be found in the documentation.

The plan for expanding Citizenship in advance of RetroPGF 3 will be published in the next few weeks! A full list of Citizens will also be published as part of RetroPGF 3, and Citizenships will be publicly recorded as attestations on OP Mainnet.

2 Likes

Can you give me an example of when this would be applicable?

It’s ultimately up to the Token House how this proposal type is utilized! By way of example, this could include:

  • a proposal to use ETH for up-front grantmaking
  • a proposal to use ETH for IRL events
  • proposals to use ETH for any other purpose that supports the business interests of the Collective

What will be the selection process and criteria for the Security Council?

More details on Security Council rollout will be shared in the next few weeks. We’re targeting Cycle #15 Review Period.

3 Likes

A short discussion on mentioned deviation in Allocation of Protocol Revenue: Token House participants hold $OP and delegate or vote with $OP tokens with the expectation that $OP’s future value will be higher than it is now. Sustainability of $OP is also important for a robust Token House that is not vulnerable to economic attacks.

Today, RPGF is justified as growing the OP ecosystem, and thus worth the investment. However, as we move to the “value” stage, this will no longer be the case. I see a gradual ramp from RPGF to dividends for $OP holders over time making sense, which has been a tried-and-tested method for cooperatives and public companies. E.g. Year 2, 100% RPGF, 0% OP holders, year 3 90% RPGF, 10% OP holders […] Year 10, 50% RPGF, 50% OP holders, and so on.

That aside, I think there’s something missing from the ideal governance - rights of users. People who don’t want to invest in $OP, don’t want to be citizens, are just using the chains anonymously. I believe they should have some veto rights for changes that’ll impact network security - i.e. sequencer selection, protocol upgrades, and security council. How this can be organised is an open question. Some inspiration can be taken from Lido’s proposed dual-governance system, of course for them it’s simpler - stETH holders are their users. It’s more difficult for a smart contract chain, but possible through oracles to determine the value of an account, or some novel mechanism. This third body would only have limited veto rights, or perhaps something simple like extending timelock duration.

PS: Although, efficient mass exit/migration solutions could also be sufficient for this.

8 Likes

hence the attestation station.

It’s not going to help users of the chain who prefer not to have an identity attached - which historically is most users of public blockchains.

3 Likes

Great post @bobby, appreciate to see a more clear path for the governance and protocol itself.

Curious about what the selection process for the new director would be like or the next steps if this situation were to happen.

Is it some sort of parameterizable requirement that the governance set rent fees for other OP chains (within the Superchain right)?

Also, I want to echo a great idea saw in the last OP governance call: a missing piece about having more clarity in the future of governance is also to have the possibility of having more clarity on legal aspects related to the foundation (and thus, the OP token status, usage, etc.) that can be of interest to governance members. I suggest a brief talk and AMA with questions made by delegates and so on.

1 Like

I think there’s something missing from the ideal governance - rights of users. People who don’t want to invest in $OP, don’t want to be citizens, are just using the chains anonymously. I believe they should have some veto rights for changes that’ll impact network security - i.e. sequencer selection, protocol upgrades, and security council. How this can be organised is an open question.

In the long term, every human community member of Optimism could conceivably be a Citizen! While today Citizenship is a big responsibility and a small set of participants, our hope is that it evolves to represent exactly what you’re calling for: a way for the human users of Optimism to express their will in governance.

There’s a lot we need in order to get there, including more scalable RetroPGF rounds, better Citizenship distribution, and a more robust identity layer – but in the long term, the Citizens’ House should represent the will of the people.

4 Likes

Lots more info about this coming soon, including today’s announcement on the Law of Chains:

I suggest a brief talk and AMA with questions made by delegates and so on.

Good suggestion! I will chat with the team. It could be helpful to gather some questions from the community ahead of time if you’re up for it :slight_smile:

3 Likes

The Citizens’ House will represent the will of the citizens. Identity is a highly subjective matter, and most users will not want to participate, unless there was a breakthrough in representing subjective outputs in an objective format (that blockchains are). Even if there was, that’s highly speculative and may take decades - we need to cater to the rights of users as soon as possible. I think a mass exit solution might do the trick though.

2 Likes

As long as the two hauses work together , it is perfect for the future of optimism governance :rocket:

Just wanted to share some thoughts about the leader election mechanism that Optimism seems to go forward with (from my understanding).
To set the stage for this discussion, will first delve into some possibilities for the decentralization process of the sequencer in general before elaborating on some specific cases :

Proof of stake

This first mechanism is a familiar concept across different Layer 1 solutions. To gain entry into the sequencer set, there’s a requirement to stake the layer 2 native token. The staking process can be realized through an Ethereum Layer 1 smart contract or directly within the rollup. The odds of sequencing a block are contingent on the volume of tokens staked.

In the realm of PoS, we acknowledge its merits as a vehicle for permissionless entry, with negative incentives such as slashing and inactivity leaks serving as deterrents to manipulation, thereby enhancing its resilience against Sybil attacks.

On the other hand, staking on Optimism may inadvertently give rise to liquid staking protocols, potentially leading to a single liquid staking provider accruing the majority of the stakes. In such a scenario, the rollup governance could become contingent on the governance token of the liquid staking provider, rather than the rollup governance token. Additionally, staking could also instigate certain regulatory apprehensions, particularly within the United States context.

Under the Proof of Stake setup, note that a rollup could opt for consensus before Layer 1, providing a stronger pre-confirmation guarantee (if decentralized) but it is still dependent on Ethereum consensus to achieve hard finality.

Based rollup

In this case, sequencing falls under the responsibility of Ethereum Layer 1. This implies that Ethereum proposers could incorporate rollup blocks within their Layer 1 blocks. This alternative directly inherits Ethereum’s decentralization properties, ensuring (without delay) liveness and censorship resistance. In contrast, the challenge here is that users typically prefer fast pre-confirmation.
With based sequencing, you remain dependent on Layer 1 block time, stifling innovation in the MEV landscape for Optimism, for instance, impeding the implementation of private mempools.

Based on governance

We then explore the concept of rollups predicated on governance. In this context, there’s no obligation to commit a significant amount of capital to be an operator in the system. Rather, it’s the rollup governance that elects the sequencers. This seems to be the direction that the Optimism collective is veering towards, and I intend to elaborate further on this choice.
Specifically, for Optimism, defined governance mechanisms are apparent, such as the selection of delegates from the token house and their capacity to advocate for upgrades. It’s important to bear in mind that members of the citizen house retain the right to veto. I presume that the sequencer election process will be governed by the Optimism collective’s governance, prompting an exploration of the potential tradeoffs:

The first pertains to permissionlessness. The need to be elected by the governance reduces the permissionless aspect of the protocol. But it raises the question: does Optimism need to have the same number of operators as Ethereum currently does (which is somewhat related to Proof of Stake)? Ethereum needs to be an entirely credibly neutral system, justifying permissionless access for joining or expanding the validator set. In my personal opinion, Optimism doesn’t have this obligation, it already inherits the security and decentralization of the base layer. It compensates Ethereum specifically for this purpose by paying for it which means that the tradeoffs are not the same and it is not necessarily needed to “copy” Ethereum. Moreover, the act of controlling access to the sequencer set could empower the Optimism governance to enforce certain behaviours that can’t be programmed client-side, such as particular types of transaction ordering, thereby eventually preventing sandwich attacks. It’s worth noting that initiatives like Flashbot’s Suave Chain, which could function as a block builder for different rollups, could eventually remedy this, although it’s not quite there yet.

The second consideration revolves around commitment. Unlike Proof of Stake, there wouldn’t be an opportunity to slash sequencers for any malevolent behaviours. However, it would be possible to get rid of them via a governance upgrade. Is this problematic? I’m not entirely sure, provided that the negative incentives are transparent and universally understood (e.g forfeiting all future revenue). It’s important to highlight that sequencers have a shorter timeframe and less power than Ethereum validators to execute a reorg. Once they submit data to Ethereum, there’s no avenue for them to reorg. Ethereum would have to revert to reorg the blocks, so sequencers can only revert on pre-confirmations prior to the data root being posted to the base layer.

The third aspect relates to tokenomics. A notable advantage is that capital doesn’t need to be locked at all, there’s no need to issue a liquid staking token. The token can remain in its native form. The value capture mechanism will likely be up to the governance (e.g allocating a portion to the treasury, retro pgf, burn). Furthermore, the payment to the sequencers could also be subject to the governance process.
Ultimately, the leader election mechanism that we obtain is a separation of capital and labor because that’s where the economic incentives lie, alongside a governance mechanism aligned with Optimism that manages the delegate selection and, consequently, the operator selection.
In reality, it doesn’t make much sense to give up a) value and b) the governance process on the operator side to another liquid staking protocol, specially when there’s an opinionated and active governance in place, which I believe is the case for the Optimism collective.

Suggestions

The delegation process in a Proof of Stake (DPoS) network is worth comparing to a governance-based network. What I mean is, when you delegate your funds in a governance-based structure, your personal funds aren’t at risk, so you avoid the potential risk of seeing users delegating their funds only to the “top” validators of the protocol. Ultimately, the delegation process is dependent on the governance that votes for the collective interest which is very different, an example of it would be electing a set of sequencers based on distributed geographical position. In my humble opinion, this would be challenging to corrupt.

It might also be beneficial to introduce a mechanism for the rotation of sequencers (e.g in a round-robin style) to ensure that each sequencer elected by the governance has an equal opportunity to order transactions and propose a block. Additionally, incentivizing honest behaviour by requiring sequencers to post a bond when they are elected could be interesting.

It would also be worth exploring the implementation of backup mechanisms in the event of a liveness failure (if the elected sequencer goes offline and a certain number of blocks are missed, other sequencers also previously elected by the governance could step in, even if it wasn’t initially their turn).
Furthermore, a reputational mechanism could potentially be beneficial to minimize the risk of this scenario recurring.

Concerning censorship resistance, forcing transactions directly through Ethereum Layer 1 doesn’t seem to be the endgame since it might be excessively costly for certain users. Therefore, having a sufficiently diversified sequencer set can help prevent this necessity. In the grand scheme of things, resorting to Ethereum guarantees around censorship resistance or even liveness will generally be slower since Layer 2 block times are typically faster than Layer 1 block times. Being censored even for a small amount of time might be too long for a user, and the sequencer set should be robust enough to prevent this.

I’m not implying that everything is perfect, and governance is a complex issue to contend with, but “enshrining” this process provides complete control to minimize negative externalities for Optimism.

It’s by no means exhaustive & certainly imperfect but looking forward to hear some of your thoughts on this topic as well. Thanks!

3 Likes