A detailed explanation of the "On-chain Court" proposed by zkSync: Is it contradictory to Vitalik's "don't let Ethereum consensus overload"?
Original Author: flowie, ChainCatcher
Most Layer 2 protocols have a "backdoor" for developers to modify contracts when necessary. For example, in the event of a major vulnerability attack, they generally need to freeze contracts through a "security committee" managed by multisig or other types of "human intervention" to limit losses. However, this "backdoor" mechanism, while ensuring the ability to respond to vulnerability crises, also leaves a centralized vulnerability that can be exploited for malicious purposes, which has been widely criticized.
To address the centralized risk of Layer 2 protocols during emergency upgrades, Alex Glochowski, co-founder of zkSync, recently proposed a new governance model called the "on-chain multi-layer court system," which simulates the real-world judicial system. The release of this new governance model quickly sparked heated debates within the crypto community: Can the "on-chain multi-layer court system" solve the centralized vulnerability of Layer 2 protocols during emergency upgrades? How feasible is it?
Handing the final verdict of on-chain disputes to the "Ethereum Supreme Court"
In order to address security issues, Ethereum has used the mechanism of forking to upgrade in the past, allowing users to independently choose to participate in forked branches that they consider to be correct and compliant. However, for Layer 2, this decentralized forking upgrade mechanism is not applicable because it cannot transfer underlying native assets such as ETH through forking.
Therefore, Layer 2 protocols have been trapped in a dilemma when dealing with security and other emergencies, where security and centralization are almost mutually exclusive. Abandoning the "backdoor" means high security risks, but keeping the "backdoor" also poses "centralization" vulnerabilities. Alex Glochowski, co-founder of zkSync, mentioned on Twitter that using a multisig-managed "security committee" can alleviate the problem by freezing funds, but it cannot solve the problem. They need token governance approval for emergency upgrades, but most malicious stakers with insufficient collateral could potentially carry out a malicious takeover of the upgrade and steal all assets. "Any government multisig granted emergency upgrade rights will bring different types of regulatory and security risks."
Therefore, in this dilemma faced by Layer 2, Alex proposed a new governance model, which is to establish an "on-chain multi-layer court" governance and arbitration system similar to the real-world judicial system.
For the higher court, it can cancel emergency upgrades and not implement any other actions. Each court must also designate a higher-level court that can appeal any decision until it reaches the Supreme Court of Ethereum. The decision of this smart contract can only be determined by a soft fork on the L1 layer of technology, and different courts will be composed in a fully decentralized manner with different members, prices, and reputations.
Regarding the "multi-tier on-chain courts" governance and adjudication model proposed by Alex, crypto analyst Haotian (@tme l0 211) further analogizes it as follows: "Special adjudication contracts on Layer 2 can be seen as county-level courts, adjudication contracts on the Ethereum mainnet layer can be seen as city-level and provincial-level courts, and the Supreme Court of Ethereum is the highest level of authority on the mainnet." So, during the appeal period, if anyone is dissatisfied with the ruling of the county-level court (special adjudication contracts on Layer 2), they can apply for a second and third appeal to the city-level and provincial-level courts (Ethereum mainnet layer) by paying a certain amount of bail, and even request the final appeal to the national Supreme Court (the highest level of authority on the Ethereum mainnet).
In other words, Alex hopes to use a more decentralized L1 layer to oversee the adjudication rights of Layer 2 in the case of emergency upgrades, so that disputes where Layer 2 is maliciously taken over and all assets are stolen can be appealed and enforced, ultimately achieving a relatively uncontroversial outcome.
But this speculative governance model is still in its early stages, and there are still many aspects of feasibility and potential risks that need to be further explored. In Alex's view, if anyone can deploy such contracts and there is no guarantee that users will bother with soft forks, then this idea needs to form social consensus and deploy a "standard" court appeals instance, which is very expensive to use and will only govern cases that are genuinely special. For example, protocols like Uniswap, major Layer 2 protocols, and other DeFi protocols with systemic risks that need to be addressed.Layer 0. Cases of concern to the social consensus layer.
Alex believes that the most important function of this "multi-layer on-chain court" system is to protect protocols from external political interference. It will serve as a powerful deterrent mechanism and enhance the role of Ethereum as a powerful network nation.
Contradiction between "multi-layer on-chain court" and Vitalik's "don't overload Ethereum consensus"?
The governance model of the "multi-layer on-chain court" proposed by Alex has sparked intense discussions as an innovative governance mechanism. Currently, many users and technologists in the crypto community are generally pessimistic about the effectiveness and feasibility of this "multi-layer on-chain court".
In a comment on Alex's Twitter, it was mentioned that there have been proposals for using L1 forks to recover vulnerabilities in L2. This was also mentioned in Vitalik's article "Don't overload Ethereum consensus" published in May this year, but Vitalik rejected such proposals.
Why shouldn't Ethereum consensus be overloaded? Vitalik provided a typical scenario where the final decision may be limited to parameter considerations at the level of ETH/USD crypto tokens, which may seem feasible. However, over time, other indices are added: ETH/EUR, ETH/CNY, and eventually the exchange rates of all G20 countries.
For instance, in the year 2034, Brazil experiences an unexpected severe political crisis, leading to election disputes. Different countries and political parties have differing opinions on this event. At this point, Brazil already has a CBDC and has split into two forks due to geopolitical differences. During the oracle voting, Ethereum stakers also diverge in their opinions and eventually propose a hard fork of the chain, causing a split in the community. However, Vitalik emphasizes that Ethereum was originally created as a globally permissionless platform free from the influence of nations and geopolitics, but it ends up being divided due to a severe internal issue in any of the G20 member countries.
In the crypto community, users such as @bitflag 123 also has similar concerns. He believes that the social consensus of blockchain is very fragile, and every fork will cause huge damage. When there is a possibility of forcing ETH to fork, it makes it difficult for small projects to compete on a level playing field, while large projects become too big to fail and gain a competitive advantage.
But regarding concerns about Ethereum splitting due to consensus overload, Alex replied that he will publicly challenge this view. He mentioned that his "on-chain multi-layer court" governance conjecture was inspired by Vitalik's perspective on "convex and concave" worldviews. He believes that adopting 100% remote governance methods is convex, and he advocates for a concave policy that involves minimum necessary intervention. This means having a non-invasive veto mechanism that pre-agrees on soft forks as governance decisions, and only using it in extreme emergencies and for protocols with systemic significance. Otherwise, one is bound to choose between two extremes: code is law, where bugs lead to death; or code is not important, and "backdoors" are allowed for response, but everything ultimately becomes centralized. Neither of these choices can serve as the foundation of a decentralized value internet.
Apart from Vitalik's suggestion to avoid Ethereum consensus overload, many users in the crypto community also believe that this governance solution faces significant obstacles in terms of feasibility. Haotian (@tme l0 211) mentioned that although the arbitration process is conducted on-chain, it is still a complex business logic that relies on time-locks and racing against vulnerability crises.
ColliderVC partner @0x idanlevin pointed out that the contradiction lies in the time-consuming nature of decision-making consultation and the need for protocols to stop-loss as soon as possible. "If an error is found and the protocol is suspended (assuming a circuit breaker or some kind of pause mechanism is used), how long does it take for the court system to reach a consensus on the fork, especially if there are several layers of upgrades? The protocol remains in a halted state with no activity or certainty for users until the solution is accepted. Protecting upgrades takes priority over centralized risks."
Web3 security analyst @0x Arhat provides a more detailed analysis, pointing out that for this on-chain multi-layer court system to work, it needs to consider the possibility of consensus, as well as the cost of appeals, dispute parameters, governance tokens, and other contract details in multiple dimensions.
For example, to make these courts function effectively, there may need to be coordination among influential individuals/groups to determine "normative" contracts and procedures. In order for this system to function properly, these considerations must all be taken into account while ensuring fairness and transparency.
The impact of the latter contradicts the spirit of the permissionless system. However, if not done this way, the Ethereum community needs to voluntarily unite and recognize specific supreme court contracts as "official" contracts, which is a very difficult process to achieve. Because there is usually disagreement on how to design or who should control it, such as whether it should be managed by DAO, smart contracts, multiple signatures of trusted developers, specific token holders, or whales, etc. If there is no social consensus, there is a risk of forming factions around competing implementations. And it seems unlikely to reach a convincing agreement on a court system without central authority.
In addition, L1 soft forks are not scalable. Given the practical coordination challenges, it is doubtful whether it is practical to rely on recurring soft forks to resolve disputes over specific protocols. The cost will also stimulate avoidance behavior, "the high cost of appeals means that the protocol may bypass the court whenever possible, thereby undermining adoption." Therefore, there needs to be a common standard for filing appeals, presenting evidence, making judgments, and other basic functions, but the details of appeal costs, dispute parameters, governance token interactions, etc., may need to be customized, which can be very tricky in practice. For example, how should appeal costs be determined? If they are too cheap, it will lead to a large number of appeals burdening the Ethereum system, but if they are too expensive, many valuable appeal cases will avoid filing.
Although currently there is a general pessimism about the feasibility of the new on-chain multi-layer court governance model, Alex stated that zkSync will provide funding for further research on this model and has attracted some developers to actively participate.
Regardless of how this new governance model will develop, the security governance dilemma of Layer 2 and most DeFi protocols in case of emergencies remains a problem worth continuous exploration and overcoming.
