Vitalik Buterin: The future will be multi-chain, pessimistic about cross-chain applications

Vitalik Buterin

2022-01-08 09:43

Read this article in 8 Minutes

AI summary

View the summary

Vitalik Buterin

原文编译： Mary Liu，比推 BitpushNews

On January 8, Ethereum co-founder Vitalik Buterin wrote on Reddit that the future will be "multi-chain" but not "cross-chain" because there are fundamental limits to the security of Bridges that cross multiple "sovereign zones."

God points out that an area of sovereignty is crossed. Apps can still, and that's not a limitation on the "modular blockchain" vision: "You can't just choose a separate data layer and a security layer, your data layer has to be a security layer as well."

The following is the transcript:

The basic security constraints of the bridge are actually a key reason, and while I'm optimistic about a multi-chain blockchain ecosystem (there are indeed independent communities with different values, and it's better for them to disagree than for everyone to compete for influence over the same thing), I'm pessimistic about cross-chain apps.

To understand why Bridges have these limitations, we need to look at how various combinations of blockchains and Bridges can survive a 51% attack. A lot of people have the mentality that "if there's a 51% attack on the blockchain, everything will collapse, so we need to go all out to prevent 51% attacks from happening at any one time". I really don't agree with that way of thinking; In fact, even after a 51% attack, blockchain maintains many safeguards, and it is important to retain them.

For example, suppose you have 100 ETH on Ethereum, and Ethereum is under 51% attack, so some transactions are reviewed and/or resumed. No matter what happens, you still have your 100 ETH. Even a 51% attacker cannot take your ETH block, because such block will violate protocol rules and will be rejected by the network. Even if 99% of the computing power or equity wants to take away your ETH, each person running the node will only follow the remaining 1% of the chain, because only that 1% of the block follows the protocol rules. To put it more simply, if you have an application on Ethereum, then a 51% attack may review or restore it for a while, but will eventually remain in a consistent state. If you have 100 ETH and sell it on Uniswap for 320,000 DAI, even if the blockchain is attacked arbitrarily crazy, you still end up with a sensible outcome -- either keep your 100 ETH or get your 320,000 DAI. An outcome of neither (or, in some cases, both) would violate the rules of the agreement and therefore would not be acceptable.

Now imagine if you moved 100 ETH to Solana Some bridge on blockchain to get 100 Solana-weth, and then what happens when Ethereum gets a 51% attack. The attacker puts a bunch of his own ETH in Solana-weth and then immediately resumes the transaction on Ethereum as soon as Solana confirms it. The Solana-weth contract is no longer fully supported, and maybe your 100 Solana-weth is now only worth 60 ETH. Even if there were a perfect ZK-SnARk-based bridge to fully verify consensus, it would still be easy to steal through such a 51% attack.

Therefore, it is always safer to own ethereum native assets on Ethereum or Solana native assets on Solana than to own Ethereum native assets on Solana or Solana native assets on Ethereum. In this case, "Ethereum" refers not only to the base chain, but also to any appropriate L2 built on top of it. If Ethereum is attacked by 51% and recovers, Arbitrum and Optimism will also recover, so the "cross-rollup" application that maintains state on Arbitrum and Optimism is guaranteed to remain consistent even if Ethereum is attacked by 51%. If Ethereum is not attacked by 51%, there is no way to attack 51% on Arbitrum and Optimism respectively. Therefore, holding the Token of the packaged version issued in Optimism on Arbitrum is still completely safe.

The problem gets worse when you operate on more than two chains. If you have 100 chains, you end up with many interdependent DApps between them, and a 51% attack on even one chain can cause systemic contagion, threatening the economy of the entire ecosystem. This is why I think interdependent regions may be closely related to sovereign regions (hence, many Ethereum ecosystem applications are closely linked to each other, many Avax ecosystem applications are closely linked to each other, etc., but not Ethereum and Avax ecosystem applications are closely linked to each other).

Incidentally, this is why Rollup can't just "go use another data layer." If Rollup stores its data on Celestia or BCH or wherever, but handles assets on Ethereum, if that layer gets 51% hit, you're screwed. The 51% protection provided by the DAS on Celestia doesn't actually help you because the Ethereum network doesn't read the DAS; It will read data from Bridges, which are vulnerable to 51% attacks. To be a Rollup that uses Ethereum native assets to provide security for your application, you must use the Ethereum data layer (and for any other ecosystem).

I certainly don't want these problems to come up right away. Attacking even a single chain is difficult and costly. However, the more cross-chain Bridges and applications are used, the worse the problem becomes. No one is going to do a 51% hack on Ethereum just to steal 100 Solana-weth (or, for that matter, 51% hack on Solana just to steal 100 Ethereum-WSOL). But if there were 10 million ETH or SOL in the bridge, the incentive to attack would be higher, and large mining pools might be well coordinated to attack. Thus, cross-chain activities have an anti-network effect: they occur infrequently but safely, but the more they occur, the greater the risk.