Whether it is technical consensus or social consensus, it is ultimately a force that constrains evil.

Author: Haotian

The security risks of Blast's multi-signature are still being criticized, and even jokes about the five multi-signatures belonging to Tieshun have spread like wildfire, but it has not affected the continuous rise of @Blast_L2TVL. Rational cognition drives everyone to criticize Blast, but the instinct to seek profit cannot resist embracing Blast.

Next, setting aside the specific multi-signature security issues of Blast, from the perspective of the layer2 industry, let's discuss why it is so difficult for layer2 technology to be decentralized.

First, clarify the decentralization of the so-called technical components of layer2, including: Sequencer, Prover, Validator, mainnet Rollup Contract, and other key components distributed across the mainnet and layer2. Most people focus on the operation of Sequencer as the core. In addition, some differences lie in the degree to which Data Availability depends on the Ethereum mainnet and the equivalence of the EVM mainnet.

The current status of decentralization of core technical components in various layer2 projects:

• Arbitrum is collaborating with Espresso Systerm to explore the possibility of decentralized Sequencer, but it is still in the exploratory stage;
• Starknet's Prover system has achieved permissionless decentralization, but decentralized messaging for Sequencer has not been clearly defined;
• Optimism has rationalized the centralization of Sequencer and introduced the OP Stack strategy, attempting to use a new shared Sequencer security governance committee to decentralize centralization permissions, and to compensate for the shortcomings of technical consensus with a curve of social consensus;
• zkSync has long shown no signs of open sourcing and decentralization of core components, and has introduced the ZK Stack multi-application chain strategy. According to the official statement, zkSync's retreat to become a model application chain under the Stack strategy is also an evasion of the sense of powerlessness in decentralized technology.

It is not difficult to see that the four major layer2 projects have their own considerations regarding the decentralization of core technical components. Some are still telling the story of technical decentralization, while others have attempted to use the Stack strategy to compensate for the shortcomings of pure technical decentralization. In short, pure technical decentralization of layer2 is very difficult. Why?

1) The Sequencer of layer2 generally uses a centralized EOA address, which can save the cost of interacting with the mainnet. Projects like Optimism and zkSync both use EOA addresses, which are directly and simply controlled by private keys, making them more flexible to operate. In addition, compared to the complex logic and functionality of smart contracts, the attack surface of EOA structures is smaller.

Most importantly, these EOA addresses can reduce costs when interacting with mainnet contracts. However, the most effective way to manage private keys is through strong centralization. If private keys were to be managed in a decentralized manner, it would actually increase the system's attack surface.

2) The Sequencer is responsible for collecting fees from layer2 revenue. Typically, the fees collected by the Sequencer, after deducting the costs required for batch transactions to the mainnet, constitute the gross income of layer2.

This makes the controlling party of the Sequencer unwilling to easily decentralize the rights. Once it becomes too decentralized, new problems will inevitably arise in the core incentive mechanism of fee collection, such as profit distribution.

3) The technical threshold for the Prover system to generate and verify proofs is high. Currently, there is little innovation in the ZK-Rollup ecosystem, and a major reason for this is the significant threshold in data structure adaptation and ZK circuit. This is especially true for decentralized Validators. Once the Validator nodes become too decentralized, there will also be stability challenges in processing and verifying proofs.

4) OP-Rollup rarely faces actual battle-tested challenges, precisely because a centralized Sequencer naturally tends towards optimistic zero challenges. To some extent, a seemingly centralized Sequencer is a soft spot, but in reality, it has become another security mechanism.

5) If a security failure occurs in layer2, the Sequencer can forcibly freeze and control the outflow of assets. At worst, a low-cost hard fork can be carried out at layer2. But what if the Sequencer is attacked? A large amount of funds would be withdrawn to the mainnet, and the mainnet Rollup contract implementing upgradeable multi-signature governance would be another layer of insurance. This is because layer2 cannot rely on mainnet-level hard forks.

To put it bluntly, the mainnet's use of multi-signature governance is only to provide insurance for the attack on the layer2 Sequencer. Discussing who is on the multi-signature list and whether they have prestige is fundamentally meaningless.

In short, the reason why we rarely hear about layer2 rug pulls due to excessive centralization is mainly because the overall technical threshold of layer2 is high, the past reputation of founding teams, the endorsement of investment institutions, and the built-in special security mechanisms (challenge mechanism, DA), all make operating layer2 a long-term sustainable business. Especially, watching the ecosystem rise one by one and continuously collecting revenue is the end game of layer2.

If a layer2 project does not mention the ecosystem, does not discuss the technology, but in the name of "everyone is the same," only talks about airdrop expectations, apart from the rug risk, the first question should be whether it is really layer2?

The current development of decentralization of layer2 technology is not ideal. Perhaps the original doctrine of decentralization does not even exist in the field of layer2? In this regard, the Stack strategy's ethereal multi-chain sharing component strategy may be the only solution to remove the excessive centralization rights in layer2 in the long run.

Fundamentally, whether it is technical consensus or social consensus, it is ultimately a force that constrains evil.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。