"Rollup on Rollup": Account abstraction solutions in the Rollup era

Written by: The Dark Side of the Moon, PANews

After the L2 technology dynamics presented "short-term OP, long-term ZK," Vitalik's thoughts began to lean towards Account Abstraction technology. At the recent Ethereum Bogotá Developer Conference, he also praised the Soul wallet, which supports Account Abstraction technology, attracting attention.

Account Abstraction can reduce the data of transactions from the source, and combined with Rollup solutions, it can serve as an integrated "data slimming" solution, significantly reducing Gas Fee consumption and usage, leading to an easy-to-use and inexpensive Ethereum network.

On October 3, Vitalik supplemented the content of the EIP-4337 Account Abstraction proposal, adding the data aggregation part of the BLS signature algorithm, allowing builders and batch validators to aggregate signatures to reduce the data broadcasted to the mainnet.

Additionally, StarkNet proposed detailed designs regarding Account Abstraction in its new version update. The DoS crisis that plagued Account Abstraction has also been resolved with the increased use of more ZK-based Rollups. The security of ZK solutions has not received the same level of attention as data compression features, which has been a long-ignored characteristic, and is expected to be fully showcased in the competition over the Account Abstraction route.

Thus, the curtain has essentially been raised on the ZK+AA on-chain data solution, giving birth to a prototype of a layered solution that is distinct from any existing public chains, which will have a profound impact on Ethereum.

Change of Mindset: Non-Consensus Layer Account Abstraction

In this update, the Account Abstraction model is designed through an EIP proposal. Initially, the intention was to change the design of Ethereum's account addresses, unifying EOA accounts and smart contract addresses, which was also the original design plan for Ethereum's account addresses.

In proposals prior to 4337, such as EIP-86/208, EIP-859, and EIP-2938, these proposals shared some common characteristics, primarily based on transforming the Ethereum mainnet. In an environment where Ethereum L2 solutions have not been widely adopted, recklessly making "bone-breaking" operations on the mainnet poses too great a risk, effectively placing the priority of Account Abstraction after L2.

This is not the optimal solution from an engineering perspective, but adhering to the principle of "it's not unusable," the practical implementation of Account Abstraction needs to wait for L2 solutions, especially the practical application of the ZK route.

Essentially, the functionalities that Account Abstraction aims to achieve can currently be realized through EOA accounts + smart contracts, without affecting the Ethereum mainnet. The community desires an "improvement," not a complete overhaul.

Ultimately, ERC-4337 received relatively widespread recognition, no longer making changes to the L1 mainnet as a consensus layer, but empowering smart contracts on it to achieve Account Abstraction in a gradual manner.

Smart Contracts: The Unified ERC-4337 Standard

As mentioned earlier, there are two types of addresses on Ethereum: EOA external account addresses and smart contract addresses. EOA controls the interaction between "humans" and Ethereum, requiring prerequisites such as private keys, mnemonics, and ETH (Gas Fee); while smart contract addresses are pre-written functions that execute specific tasks under certain conditions.

The two work together to create the richest Ethereum ecosystem, and concepts like mnemonics, wallets, private keys, and smart contracts are deeply ingrained. However, this is not the optimal solution for Ethereum; maintaining two sets of address types reduces the operational efficiency of the mainnet, as it requires additional overhead to determine what type of address an address belongs to, not to mention that complex operations incur expensive Gas Fees.

Therefore, since 2015, there has been a continuous search for merging the two. The current 4337 proposal is the closest to a successful design, essentially using a smart contract solution or endowing Account Abstraction with a higher-level "intelligence" to unify the two, rather than modifying the mainnet at the consensus layer to seek a balance between security and efficiency.

Users are expected to send off-chain information called User Operations, which are collected by block builders and packaged into a batch transaction. Builders are responsible for filtering these operations to ensure they only accept fee-paying operations. There is a separate Mempool for User Operations, and nodes connected to this Mempool perform ERC-4337 specific validations to ensure that User Operations guarantee fee payment before forwarding.

The benefits of this approach include:

• More intelligent. The logical characteristics of smart contracts will enable more complex operations;

• More secure. Wallets can have built-in multi-signature features and social recovery functions, eliminating the need for mnemonics.

• More flexible. For example, it can enable sponsored transactions by others, avoiding the predicament of having no ETH to pay for Gas.

Essentially, the Account Abstraction designed in ERC-4337 "delegates" smart contracts to ordinary accounts, functionalizing some complex logic, allowing users to experience greater convenience at the front end without needing to deeply understand the underlying logic.

However, the logical complexity of smart contracts can also lead to additional account overhead, which is why continuous technical optimization is needed, along with Rollup solutions to achieve overall speed and cost reduction. Together, they form the main direction for improvement in the post-PoS era, with priority only after the ultimate solution, DankSharding.

A Perfect Match: The Convergence of Account Abstraction and Rollup

Before the implementation of Account Abstraction, let's summarize the current characteristics of Account Abstraction and Rollup to find points of convergence:

1. Account Abstraction: Confidential, intelligent, universal, but Gas Fees are high, primarily due to the increased data from multiple steps, which is exactly what ZK-Rollup excels at reducing;

2. Rich functionality: Multi-signature wallets ensure security, and permission segmentation facilitates governance, but the resulting logic can become more complex, and security vulnerabilities may expand accordingly;

3. Not yet explosive: Rollup is still in the pre-ZK era, and the current Account Abstraction solutions are based on the premise of large-scale L2 practical applications, remaining a pre-research technology;

4. Future alignment: Once L2 gradually takes shape, the consensus layer will no longer change, and smart contracts will have more room for application, ensuring the security of Account Abstraction.

While Account Abstraction can enhance the intelligence of addresses, it is currently not suitable for direct user promotion due to its high complexity, which would directly raise Gas Fees and reduce security. For example, in sponsored transactions, if the mechanism fails, it could directly lead to a crisis in the operation of the mainnet.

The increase in Gas Fees is also detailed in the proposal, which is the direct motivation for Vitalik's recent supplement to ERC-4337. Theoretically, each step of an Account Abstraction address requires computation, needing more computational resources to ensure security without human intervention. Transactions need to include the one-time setup costs for contract addresses and proxy addresses.

Traditional EOA external account addresses do not require such "complex" transaction logic; they have already been verified by humans and set up on the Ethereum mainnet, allowing for immediate use. The Gas Fee for account transactions is about half that of Account Abstraction.

Therefore, before a more cost-effective solution is implemented, the complexity of Account Abstraction may hinder the operational speed of Ethereum, which is precisely the opportunity for ZK-Rollup, as it is better at presenting global verification results with compressed data.

The proposal content updated by Vitalik actually connects the data interfaces of Account Abstraction and Rollup, allowing L2 solutions to also handle the data generated by Account Abstraction. The improvement lies in packaging functionalities that can only be achieved step by step into batch transactions, closely aligning with the Rollup approach while still ensuring transaction validity through SNARK technology.

One can imagine that after the combination of Rollup and Account Abstraction, Ethereum will exhibit characteristics of universal data, convenient verification, and high security, without needing to modify the consensus layer, relying instead on more refined solutions.

Similar to various Rollup solutions, Account Abstraction also has the potential to trigger an ecological explosion. For instance, the largest smart contract multi-signature wallet in the EVM ecosystem, Gnosis-Safe, may find its technological moat no longer exists, and a new narrative model will promote the emergence of updated solutions.

In Layer 2 solutions, StarkNet is also attempting to integrate Account Abstraction functionality, incorporating it from the design phase, including the Account Abstraction module specified by Entry Point contracts, separating verification and execution, and adding a Nonce field to enforce transaction uniqueness at the protocol level.

Moreover, StarkNet will construct direct communication between L2 and L1 to better coordinate the state transmission mechanism between the two, allowing the advantages of Account Abstraction to be realized under the premise of compressed data costs, making it smarter and cheaper.

Conclusion

Since the first practical Account Abstraction proposal was put forward in 2015, more than seven years have passed. Against the backdrop of the gradual advancement of ZK-Rollup, truly realizing the functionality of Account Abstraction will still take a long time. However, just as the transition from PoW to PoS has been achieved, Ethereum is developing along its established path. Account Abstraction, as a significant foundational feature planned alongside PoS, Layer 2, and DankSharding, will see broader usage after the realization of PoS and ZK-rollup-based Layer 2.

With more and more Rollup-compatible Account Abstraction solutions, we may soon be able to discard mnemonics, truly achieving a Web 2-level user experience combined with the security and privacy of Web 3.