Spend an hour clarifying the account abstraction

The core advantage of ERC-4337 lies in the classification of the application layer. For abstraction, it is not difficult to modify the underlying architecture, but it is difficult to achieve compatibility with historical burdens.

Written by: Fourteen Jun

2 months ago, I received an invitation from the hacker dojo platform, only to find that in this impetuous era, there is such a platform that only focuses on the underlying technology and shares it for free. I also want to try to use the form of live broadcast to explain clearly the underlying logic of the 4337 account abstraction.

This article shares the live courseware and sorts out some core questions of the audience. (See station b for the full video.)

Full text overview

1. The Road to Account Abstraction

1.1, wallet track classification

The so-called account abstraction, what is his account?

Now there are two types of accounts, one is called EOA account and the other is called CA account. It can be said that all contract wallets except 4337 can be called a non-standard contract wallet, MPC wallet. It is also a very hot field now. A while ago, it could be said to be a street in Silicon Valley. Then at least five companies were found to be making MPC wallets, and most of them used the TSS method. It is equivalent to signing each shard individually. After signing, everyone gathers.

There is a significant difference between the MPC wallet and the account wallet. EOA and CA are more about the subject on the chain. mpc refers to your off-chain. A technical solution for private key management.

**1.2. Why abstract? **

Three types of transactions can be distinguished on Ethereum: Legacy type, EIP2718 type, and Eip1559 type. The current mainstream is the 1559 type, which aims to cut the price structure in traditional transactions.

Then there is such a classification. We can know that they have nothing to do with account abstraction, so we need to look at the role of abstraction from the account structure and transaction structure itself. The core structure is:

• nonce: anti-replay
• balance: balance accounting model
• storageRoot: the MPT tree root of the contract's Storage variable KV
• codeHash: the hash of the contract code, which is also the key to find the code bytecode

You can also see his fields from the transaction structure

• Nonce
• Throttle limit (START GAS)
• Gas Price
• To (MSG.SENDER)
• VALUE (MSG.VALUE)
• V, R,S (Signed TXwith SENDER)
• Data Bytecode

You can say anything, but there is no From address, because the actual From information is based on the ECDSA signature unsigned result, so whether it is 4337 or Native AA, you can customize the From party, which means that you can use the Control is separated from ownership and account subjects.

This is what we call abstraction now, and it not only abstracts some explicit data, but also extracts the hidden data of this transaction.

**2. Overview Why is the EIP 4337? **

Further reading: Ethereum Account Abstract 4D Research Report: Dismantling 10 Related EIP Proposals and the Seven-Year Road to Breaking the Bottleneck of Tens of Millions of Daily Active Users

** Summarizing the process of appeal, we can find that the core advantage of 4337 lies in the classification of the application layer. It is not difficult to abstract this matter and modify the underlying architecture, but it is difficult to achieve compatibility with historical burdens. **

3, System structure overview of 4337

3.1 Advantages and disadvantages of comprehensive comparison schemes

3.2, Operation Mechanism and Performance

3.3, Implementation details of the EntryPoint contract

Extended reading: Interpretation of the abstract ERC4337 review scheme of the Ethereum account (Part 1)

4. Q&A session

**1. Can the problem of high gas fees in smart contract wallets be solved? **

Part of it can be solved. Whether the gas fee is high or low needs to be compared to find out.

First of all, the composition of gas is as follows: a single eoa transfer transaction is at least 21,000 gas or about $0.22, and the smart contract wallet will require higher gas because it needs to increase the deployment contract, scheduling contract, contract deduction calculation, So it must be higher than eoa transfer.

Secondly, the solution is based on the bundled transaction of 4337. Since one transaction is at least 21000 gas, and the bundled transaction is multiple transactions sharing 21000 gas, if 10 transactions are reported together, it will be 21000+10*operation consumption, as long as the operation If the consumption *10 is greater than 21000*9, it will be cost-effective.

Finally, the core reason for the high gas is that the performance of the eth coin itself is limited, and the price of the currency is too expensive, so we have to wait for the development of l2 to bring lower-cost concurrency, so as to fundamentally solve the problem of high gas.

In short, L1 is still more suitable for storage, and L2 is the way to break the performance

**2. Can the traditional EOA wallet iterate out the social recovery, multi-signature, batch payment and other functions of the smart contract wallet? **

Some functions can be iterated out, but most cannot.

The traditional wallet is based on the ECDSA algorithm, and the public key is fixedly calculated by the private key. It cannot have the characteristics of a contract wallet. There was a proposal to change the signature algorithm. If implemented, it is possible to achieve multi-signature (or realize the current sharding key based on mpc+tss Multi-signature), as for social recovery and batch payment, it is impossible to solve through cryptography.

**3. How to understand MPC wallet and smart contract wallet, these two directions to improve user experience? **

I personally think that this is two-dimensional management. The mpc wallet manages the shard keys off the chain, while the contract wallet manages the main body on the chain and provides programmable on-chain management capabilities.

Both can be used at the same time. After the contract wallet is perfected, it can replace mpc's multi-signature and recovery changes, while mpc cannot provide the advantages of contracts

• eoa wallet: easy to register, has a long history, and can also achieve multi-signature based on chain contract wallets, etc., partial to individual users
• mpc wallet: It is naturally multi-signatured, easy to prevent asset loss, and it is newly launched in recent years, so it is used for institutional treasury management
• Smart contract wallet: Currently, the cost of use is relatively high, suitable for people with high assets, asset circulation, open and transparent transaction operations, easy to vote for dao and public execution logic

**4. From a technical point of view, will all wallets be smart contract wallets in the future? Is this a recognized trend? Will the current EOA wallet giants be eliminated or evolve into other forms? **

Yes, 4337 is also just a transition state, the final form is Native AA

**5. If there is a good enough MPC+TSS solution, users do not need mnemonics and the security is guaranteed, then do we still need a contract wallet? **

Behind this problem is the compromise between the impossible triangle, and the achievement of both security and convenience will inevitably affect the degree of decentralization. The contract wallet also has additional functions such as trusted operations, time locks, and social recovery, and can continue to be used as a DAO team treasury and other scenarios.

**6. 4337 still relies on EOA account signature, can it really bring a qualitative leap to user experience? **

At present, in the short term, it will depend on the signature of the EOA account. In the long run, it will inevitably need to manage the corresponding key of a certain signature algorithm. The complexity remains the same, and even the logic of the operation becomes more complicated.

Therefore, although there are improvements in function points such as contract-based wallets, that is, bundled transactions, the user experience itself will not have a qualitative leap.