The Archeology of Ethereum Account Structure and the Value of Account Abstraction

Before discussing the value of account abstraction (AA), it is necessary to review the Byzantine problem and the pain points of Ethereum.

On the traditional Internet, when A pays B for online shopping, two things happen: 1. A transfers the value of the money to B for value transfer; 2. A notifies the payment of one dollar B; From the perspective of information, A loses one yuan, B gains one yuan, and A will get a certain product. Only after the above three information are confirmed at the same time, an online shopping order is considered complete. How to confirm the above three processes simultaneously can be understood as a Byzantine problem.

What Bitcoin or the Byzantine problem wants to solve is to realize the atomic operation in the real world in the Internet environment or in the computer environment, that is, when the value transfer occurs, the information transfer also occurs. However, based on its concept of decentralization, Bitcoin cannot construct a complex Internet environment.

In this regard, in 2012, Vitalik Buterin, who has not yet become a god, initiated a proposal on Bitcoin Core, hoping to increase Turing completeness on the basis of Bitcoin to achieve more functions. After being ruthlessly rejected, he established the Ethereum ecosystem. Although the birth of Ethereum and smart contracts has brought the blockchain world to a new level, Ethereum running in a decentralized environment still faces the biggest pain point: the linear environment cannot perform high concurrency and complex code compilation; This is why Vitalik has been encouraging users to enter the layer 2 network, advocating contract wallets and account abstraction, and even encouraging users to hand over some privacy to the project party in exchange for social login, social recovery and other user experiences close to Web 2.0: If Ethereum does not Make these changes, and it can't live up to its vision, forever remaining as an adjunct to Bitcoin.

Core topic 1: What exactly is account abstraction?

There are two types of accounts on the Ethereum public chain, one is called external address account (EOA), which is characterized by free creation and can directly initiate transactions through signatures; the other is contract account (CA), which has its own code area and The data area is essentially a smart contract, so transactions cannot be initiated directly; because creating a contract account = creating a smart contract = executing on-chain interaction, so you need to pay* (this is V God’s previous statement about MPC wallet and CA wallet In the argument, it is mentioned that the CA wallet creation fee is a cause and effect of xx dollars)*. And when EOA sends an asset to CA, theoretically this CA can become an account abstraction (Account Abstraction) wallet.

Account abstraction (AA) wallet can be understood as a decentralized "bank". EOA can put assets in this "bank", and just notify the "bank" when transfer is required, and the "bank" will complete the transfer operation. In this way, value transfer and information transfer no longer occur simultaneously on EOA, but on an abstract account. Of course, these transactions ultimately require EOA to issue transaction instructions and sign for execution.

Core topic 2: How is the function of account abstraction realized?

With the update of the content of the EIP-4337 protocol, the hype (or deification) of the abstract concept of accounts in the capital market has reached its peak, and the most commendable ones are [social login] [social recovery] [strange device recovery] [no handling fee] Transfer] and other concepts; but in fact, the contract account (CA) itself is a decentralized product, and the operation of the transaction on the asset chain still needs to be executed through the EOA signature, so through tools such as web 2.0 information, biological information and authenticator 【Social login/recovery】Centralized service cooperation is necessary* (this is also the reason mentioned above that "V God even encourages users to give up some privacy in exchange for a user experience close to Web 2.0")*.

However, [Strange Device Recovery], a controversial function* in the EOA wallet (for example: the private key recovery service launched by the Ledger wallet)*, can be implemented in the CA wallet by writing a spare EOA in the smart contract. There is no need to hand over any sensitive information to the wallet operator, and there is no need to transfer any private information, so it is safer than the recovery function launched by the EOA wallet.

Finally, how is [transfer without handling fee] realized? First of all, everyone should pay attention to an important point: CA wallet and account abstraction (AA) wallet can only run on public chains that support EVM at present, and the main battlefield must be Ethereum. As we all know, before the implementation of the EIP-1559 proposal, the POW miners of Ethereum can freely choose an on-chain transaction to pack for free, but after the EIP-1559 proposal is passed, the base fee must be paid due to deflation and combustion, causing Ethereum There are no truly free transactions on the web. Next, Cregis Research will analyze in detail how the AA wallet [transfer without handling fee] is realized:

First, take the USDT smart contract as an example: it includes a code area and a data area. The data part can be understood as a table that records how many USDT each EOA owns, that is, the balance of each EOA. When transferring money, the balance of one account is deducted and the balance of the other account is added.

At this time, we substitute the thinking of game-fi or social-fi: the project party hopes to lower the threshold of participation, so that users can participate without holding gas fee. Although the current Ethereum EOA cannot do it, smart scientists still think of an indirect solution: Meta Tx & Gas station network (GSN), the principle is that the token trader (Zhang San) signs to execute a transaction-like structure The data (mdata), and then sent to the person who is willing to pay the gas fee (Li Si), Li Si will send the transaction containing the mdata to the smart contract compatible with GSN for execution, and pay the gas fee on the chain to complete the transfer, and the Ethereum network The initiator of the consensus node confirmation transfer is Zhang San.

The above solution is an important part of the EIP-4337 proposal. The EIP-4337 proposal is a transaction optimization scheme that does not change the Ethereum consensus layer: when a user wants to initiate a transaction, TA will only need to send a msg, but this msg is not sent to the Ethereum mainnet, so the user does not need to pay gas fee; this msg will be sent to a memory pool, where it will be authenticated according to the EIP-4337 standard, and if it is confirmed to be correct, it will be published to the Ethereum chain by the EOA of the operator, and finally the POS node will send the transaction Entry block. In addition, the EIP-4337 proposal also includes the benefits that the packager can obtain, and a series of operation standards such as how to refund if the packaging fails; PS. If the EIP-4337 proposal is added to the method of hiding the association between the operator's EOA and msg , it can also realize the function of private transaction.

EIP-4337 Proposal: No Changes to the Ethereum Consensus Layer

Mystery solved: account abstraction (AA) wallet = contract account (CA) + off-chain communication standard

After the analysis, I believe everyone understands that although the content of the EIP-4337 proposal may take 11 to 13 months to be fully completed, the account abstraction (AA) wallet is not only realized through the EIP-4337 protocol, so there are indeed The genuine AA wallet product, but the EIP-4337 protocol is likely to be the most recognized operating standard and solution.

As an established enterprise-level wallet, Cregis can certainly provide CA and GSN services: CA is optional in privatization deployments, and GSN was deployed on the TRON network in April this year, saving customers 1.93 million TRX so far; but Cregis is not in a hurry to launch the AA wallet service publicly, because the above-mentioned industry standards have not yet been implemented, and rash launch may cause future products to be incompatible with mainstream standards.

*Gnosis Safe, a well-known multi-signature wallet operator, also realizes the multi-signature function through the contract account (CA). It is also not eager to launch the AA wallet. It is speculated that there may be concerns in this regard. *

Core topic three: Is the AA wallet really perfect? What are the disadvantages?

The advantages of AA wallets are many, but they are not perfect. A week ago, God Vitalik commented on CA and MPC based EOA on his Twitter AMA. Various masters also discussed in depth on this Twitter. Here are a few objective flaws for everyone:

• The cost of creating a CA is high, and the more prosperous (more congested) the Ethereum network is, the more expensive it will be;
• The security of CA depends on the builder of the smart contract. If there is a loophole in the smart contract or the builder leaves a malicious backdoor, the security of user assets is at risk. The staking contract of the cross-chain bridge is also a CA. The theft of the cross-chain bridge and the huge amount of stolen assets that frequently hit the headlines are all risk warnings for the CA wallet;
• CA can currently only run on the EVM chain, and does not support token collection of BTC and non-EVM public chains. This is why although Cregis can provide CA, in order to take into account the diversity of enterprise-level user collections, other wallets must be provided at the same time The reason for the solution.
• Calling the CA wallet relies on the EOA signature. The EOA generated through the seed phrase needs a private key to be signed. If the private key is not kept properly, it will still be stolen, and the security logic has not been upgraded;

Cregis Research, as one of the onlookers of this popular AMA, believes that there is no absolute safe path in the blockchain world, and it is the responsibility of each user to make a balance between convenience and security; and to choose Cregis It is the responsibility of Cregis to provide users with tools that are both safe and convenient: a fully self-hosted MPC (multi-party computing) wallet + optional privatized deployment solution (including CA) + rich financial SaaS functions to start a business for Web 3.0 Providers provide a new asset collaboration management platform.