zerokn0wledge

L2Beat rollup stages explained: Stage 0, 1 and 2 criteria and who qualifies

Post732 words

Guide to L2Beat's three-stage rollup framework (training wheels) with criteria per stage, and which rollups sit where. Most are Stage 0, Arbitrum and dYdX are Stage 1, and only Fuel v1 and DeGate reach Stage 2.

Key claims
Stage 1 needs 5 external fraud proof submitters, 7-day exit window and a council of at least 8
Stage 2 requires a 30-day exit window
Only Fuel v1 and DeGate were Stage 2 on mainnet
A Guide to L2Beat's trust framework for rollups Rollups have emerged as the most promising solution to trustlessly scale Ethereum. However, the early development phases of rollups often involve centralized control mechanisms, also referred to as “training wheels”, to enable streamlined updates and bug fixes. However, for rollups to truly inherit security from the Ethereum base layer, these training wheels need to be taken off over time. @L2Beat, inspired by Vitalik Buterin’s proposals, has therefore introduced a framework to categorize rollups based on their dependence on these training wheels. The framework classifies rollups into three stages: Stage 0️⃣ - Full Training Wheels: Predominantly managed by operators but offers software for state reconstruction from data on L1 to verify state roots. Stage 1️⃣ - Limited Training Wheels: Governed by smart contracts with a Security Council in place. This stage includes a functional proof system, decentralized fraud proof submission, and allows users to exit without the need for operator involvement. Stage 2️⃣ - No Training Wheels: Entirely controlled by smart contracts with a fully permissionless fraud proving mechanism. The Security Council's powers are confined to on-chain error rectification. Historical Context Vitalik Buterin's proposed rollup milestone framework on the Ethereum Magicians’ forum in November 2022 ignited a broad discussion. This conversation highlighted the need for a detailed rollup maturity assessment framework. @l2beat revisited multiple rating systems before deciding on a simple three-stage framework, refining the definitions initially put forth by Vitalik. This structure ensures clarity, precision, and offers a comprehensive technical breakdown for deeper insights. Detailed Framework Breakdown Stage 0️⃣ Criteria: 🔹 Identification: The project must call itself as a rollup. 🔹 State Roots on L1: State roots should be posted on L1 to facilitate withdrawals. 🔹 Data Availability (DA) on L1: All data to reconstruct the L2 state should be available on L1 for transparency and auditing. 🔹 Software for State Reconstruction: Rollup node software should be accessible, enabling independent validation of state roots against provided data. Stage 1️⃣ Criteria: 🔹 Proof System: Required to determine the authenticity of the proposed state root. It must ensure all state changes are incorporated in data availability. 🔹 Fraud Proof Submission: At least 5 external actors should be allowed to submit a fraud proof. 🔹 Independent User Exit: Users should be able to exit without the operator's intervention. 🔹 Exit Window: A minimum of 7 days should be provided for users to withdraw their assets in case of unwanted network upgrades. 🔹 Security Council Formation: Must be transparent, diverse, and include at least 8 participants, with a 50% consensus threshold. Stage 2️⃣ Criteria: 🔹 Permissionless Fraud Proof System: It should be open and decentralized (in optimistic case). 🔹 Extended Exit Window: A 30-day period should be provided for users to withdraw assets in case of undesired upgrades. 🔹 Limited Security Council Intervention: The council should only intervene for on-chain detected errors, reducing trust dependence on them. Most rollups live today are still in stage 0. This includes: 🔹 @optimismFND 🔹 @BuildOnBase 🔹 @zksync Era 🔹 @Starknet 🔹 @loopringorg 🔹 @LineaBuild 🔹 @0xPolygonLabs zkEVM 🔹 @MantaNetwork Pacific 🔹 @ourZORA 🔹 @kroma_network Stage 1 rollups include: 🔹 @arbitrum 🔹 @dYdX 🔹 @zksync Lite Currently, there are only two "stage 2" rollups live on mainnet, according to the @l2beat framework. One is @fuel_network v1, which is the only optimistic rollup that has a fully permissionless fraud proving mechanism deployed. The second one is @DeGateDex, the first zk rollup to reach stage 2. Stay tuned for a thread on it next week! Last but not least, there is a number of Validiums and Optimiums (rollups that don't use Ethereum L1 for DA), which are not categorized based on this framework. This includes: 🔹 @0xMantle 🔹 @Immutable 🔹 @MetisDAO 🔹 @OfficialApeXdex 🔹 @Arbitrum Nova 🔹 @Sorare 🔹 @rhinofi It's worth noting, that Manta Pacific will also fall into this category, once they have integrated with Celestia for DA and effectively become a Celestium (Optimium with DA on Celestia). Conclusion L2Beat's trust framework is a great tool that helps users to understand the trust assumptions that underly all these emerging L2s. By breaking down the journey of rollups from a centralized to a decentralized networks into tangible stages, L2Beat provides projects with clear milestones, improving user understanding of a rollup's maturity and trustworthiness. If you want to check the individual rollups in more detail, check the @l2beat website here! 👇 🔗 Link: l2beat.com/scaling/summary

1 image on X

Open on X
zerokn0wledge avatar
Working together

Research, advisory and builds for deep-tech teams.