Delphi: Dissecting Based Rollup — New ETH Alignment

Author: Utonio, Delphi Digital, Translation: Jinse Finance xiaozou

L2s saved ETH from the loss of users, but it brought problems such as liquidity fragmentation, ecological isolation and reduced composability. Are Based rollups, shared sorting and pre-confirmation solutions? Utonio specifically analyzed each solution in his in-depth study of Based Rollups and formed the following 4 key insights.

1. Properties and limitations of Based Rollups

Simplicity: Based sorting simplifies the process and does not require additional verification or hard forks

Trusted neutrality: Use decentralized Ethereum sorter

Minimize risks: Inherit the core layer L1 security of Ethereum

Flexibility restrictions: Bound to the timing and infrastructure of the base layer, weak independence

Delay: linked to L1 times, may be offset by pre-confirmation

2. Conflict

While based rollups retain the ability to issue governance tokens and charge base fees, it conflicts with the current rollup business model, in which sorters have transaction sorting rights, can obtain MEV and earn tips.

If they were to switch to based rollups, sorters would give up this value to L1 validators and limit their income to base fees.

How easy is it for current rollups to become based rollups?

The answer is quite easy. Current rollups like Arbitrum can be converted to based rollups by abandoning their centralized sorters and allowing transactions to be sent directly to the Inbox contract on L1 to be sorted by L1 proposers.

This will eliminate the involvement of third parties while maintaining L1 latency. The protocol will allow anyone to create transaction packages, compress them, and publish them on L1, rather than limiting the sorter to publish transaction packages.

Because execution still occurs on rollups, error proofs and zero-knowledge proofs are still necessary because they are about confirming execution outputs.

3. Preconfirmation

Preconfirmation is gaining momentum and is committed to solving the main limitations of based rollups.

L1 proposers (who also act as L2 builders for based rollups) can now choose to perform additional tasks by re-staking to become pre-confirmers (preconfer), a new name for sorters that sort transactions in L2.

Users tip preconfers for transaction inclusion, and preconfers may also receive native token rewards.

Rational economic actors should choose to join, either through delegation or re-staking. The growth of clients driven by MEV proves that participants aim to maximize revenue.

Preconfers utilize L1 block times to complete transaction commitments, and can include multiple L2 blocks in a single L1 block.

Delegating preconfirmation rights introduces trust assumptions and comes with penalty conditions related to liveness and security errors. We explored different types of implementations, such as Based Espresso, designed and developed by the Espresso Systems team.

4. Based sorting pattern

The entities dedicated to based sorting are listed as follows:

Focus on Taiko

Taiko belongs to zk-EVM Type 1 and is the first based rollup project built from scratch without making any changes to the Ethereum system. The project is a permissionless project with no centralized sorter or whitelisted provers, allowing anyone to run proposer and prover nodes.

Taiko adopts a multi-proof approach, using multiple proof types (such as multiple SNARKs + multiple clients). Provers are incentivized by TKO token bonds, which can be slashed if they fail to prove blocks within the expected time.

They are leveraging the RISC Zero zkVM infrastructure to build the Taiko execution VM, which will make it possible to build ZK apps.