Partnered With:

Compare the advantages and disadvantages of Arbitrum and Optimism in the process of dispute resolution in detail

Read this article in 19 Minutes

Optimism's dispute resolution process is simpler and faster than Arbitrum's because it only needs to provide the disputed transaction through EVM.

原文标题：《 Making Sense of Rollups, Part 2: Dispute Resolution on Arbitrum and Optimism 》

原文作者： Benjamin Simon

原文编译： ETH 中文站

All Rollups follow a similar basic architecture and internal logic. Nevertheless, as we said in the first part of this mini-article series, the single difference between Optimistic Rollups and zero-knowledge Rollups (i.e., the way their review processes operate differently) accounts for a large number of other differences. Such as differences in security, availability, and EVM compatibility.

The same is true for every Rollup type. Although Arbitrum and Optimism, the mainstream Optmistic Rollups solution, have much in common, it is not just user loyalty that distinguishes them. Rather, their respective approaches to resolving disputes create performance trade-offs. Given that both schemes aim to provide Ethereum with full capacity expansion, these tradeoffs are worth discussing.

In the early days

Let's start with a brief historical background for each project. As it happens, both have some unique origin stories.

On a cold Princeton morning six and a half years ago, a group of undergrads working with Professor Ed Felten gave a talk about the project they had signed on to build: a blockchain-based arbitration system. The goal is to sidestep some of the expected scaling challenges of smart contract platforms, and the plan is to ease the computing load of traditional miners by designing a blockchain that relies on challenge and dispute resolution systems. Had it not been for Steven Goldfeder and Harry Kalodner, two ambitious PhD students who came up with the idea of building a powerful Layer2 solution based on their original concept and approached Felten with that idea, The Arbitrum system will suffer the same fate as other promising academic computing science projects. Soon thereafter, Felten, Goldfeder, and Kalodner co-founded Offchain Labs, which turned Arbitrum from an abstract idea into a concrete reality.

Optimism is preceded by Plasma. In mid-2017, Vitalik Buterin and Joseph Poon co-authored a paper proposing Plasma, an early Ethereum scaling solution. A group of core Ethereum researchers then took over the project's development and formed a nonprofit research group to build the vision. In 2019, however, Plasma's development stalled as some of its key design limitations became apparent. Karl Floersch, Jinglan Wang and Ben Jones, the three main researchers of Plasma, were not stopped by the shock and decided to study the evolution of Plasma as an Optimistic Rollup. They formed the Optimistic PBC team in early 2020.

Dispute Resolution: A brief (re) introduction

Recall that Optimistic Rollups takes a "assuming something is valid until it is falsified" approach to ensure the validity of a transaction. Optimistic Rollups processes the transaction and feeds the results back to the Ethereum network so that it can finally package the transaction into the underlying chain. In the dispute phase, if the Rollup orderer processes an invalid transaction, then everyone monitoring the state of the Rollup chain can submit a challenge. The dispute resolution process is then triggered immediately. The difference between Arbitrum and Optimism is that they have different approaches to dispute resolution (both in terms of cost and time).

A Preliminary Comparison between Arbitrum and Optimism in Dispute Resolution

The easiest way to describe the difference between the two is this: Compared to Arbitrum, Optimistic dispute resolution relies more on the Ethereum Virtual Machine (EVM). When a user submits a challenge on Optimism, the entire transaction associated with it is run through EVM. In contrast, Arbitrum uses an off-chain dispute resolution approach to reduce disputes to a single step in the transaction. The agreement then sends this single-step assertion to the EVM for final validation, rather than sending the entire transaction to the EVM. Therefore, conceptually speaking, a Optimistic dispute resolution process is much simpler than that of Arbitrum.

In the case of Arbitrum, the off-chain component of its dispute resolution process uses a recursive dichotomy algorithm. This sounds complicated, but in reality, the algorithm simply forces the "asserter" (who processes the transaction) and the "challenger" (who submits the challenge) to move back and forth to narrow the point of contention, as shown in the figure below. Interestingly, this process of resolving disputes back and forth through recursive dichotomy was originally conceived in Arbitrum in 2015.

Source: Offchain Labs Development Center

A Optimistic dispute resolution approach (running the entire transaction through EVM) is not only conceptually simpler, but also faster to run. Optimistic dispute resolution does not require multiple rounds of processing back and forth like Arbitrum does. In fact, due to the reason, the Optimism rollups is often become a "single wheel" (single round), and Arbitrum was become "rounds" (multi round). In fact, for a disputed transaction, if it's on Arbitrum, it takes longer for it to get final confirmation on Ethereum than for it to get final confirmation on Ethereum. As we discussed in the first part of this series, the speed of dispute resolution is particularly important because it determines how long it takes for users to be able to transfer their tokens from Rollup to Ethereum.

On the other hand, the advantage of Arbitrum dispute resolution is that it has lower transaction costs on the chain (such as Ethereum). Because after Arbitrum has completed the back-and-forth dispute resolution process, a small piece of code submitted to the EVM for final processing costs less gas (in most cases [1]), while reprocessing the entire transaction on the chain costs more gas.

Refactor the comparison

The basic trade-off between these two dispute resolution designs is simply between speed and on-chain cost. In practice, however, this framework is somewhat simple, as few people believe that there are two reasons for the controversy:

Due to its economic incentive/penalty design, neither Arbitrum nor Optimistic transaction processors deal with fraudulent transactions. They are forced to post a Pledged Asset/Bonds to Arbitrum or Optimism in advance, and if they process a fraudulent transaction, their assets will be forfeited.

Similarly, parties to the monitoring rollup chain are reluctant to submit false proof of fraud -- in this case, because the challenger has to pay the on-chain gas fee for the proof of fraud; In Arbitrum, because the challenger loses the dispute, the margin is forfeited.

Given that disputes are expected to be minimal, why is the structure of the dispute resolution process so important?

Although controversies are rare, Rollup must be designed with the assumption that a controversy will occur at any time. Thus, the design of the "disputed" case affects the structure of the "undisputed" case.

Since a Optimistic must be able to run each transaction through EVM when a dispute occurs, it cannot process transactions that exceed the Ethereum gas limit because these transactions cannot be verified correctly on the chain. In contrast, Arbitrum can execute transactions of any size, even though they exceed Ethereum's gas limit, as these transactions do not need to be run in bulk through EVM, but are first broken down into tiny "assertions".

It is not clear how much of a realistic limitation a certain gas limit can impose on an application. However, a more important implication of the differences in dispute resolution design is that Arbitrum can save gas consumption by reducing the frequency of checkpoints (update status root) on the chain. More specifically, Arbitrum can allocate a large number of off-chain calculations to a single update, since theoretically the status root update can include (tiny) single-step fraud proofs for all transactions contained within it. On the other hand, a optimist must perform a checkpoint on the chain after each transaction, significantly increasing the footprint on its chain.

Overall, Arbitrum should save more gas than Optimism -- and therefore be cheaper for the user -- and not only in rare cases of controversy, but also in cases where there is no controversy.

Dispute Resolution and Potential Attack Vectors

A final point worth discussing about the two different dispute resolution processes is how resistant each design is to potential attacks. In the previous article, we talked about financial incentives to stop spamming attacks. More specifically, both Optimism and Arbitrum validators are reluctant to submit ineffective challenges.

But what if there are some malicious attackers who don't mind the economic cost of spamming Rollup? In other words, what if a person or entity is so committed to slowing down the progress of an Optimistic Rollup (and they are willing to do so, even if the attack means repeatedly paying for ineffective challenges)?

As mentioned above, Optimism's dispute resolution process is simpler and faster than Arbitrum's because it only needs to provide the disputed transaction through EVM. This speed is a factor of Optimism here, as disputes can be resolved quickly without stopping the rollup chain from developing.

The concern is the "multi-round" dispute resolution approach used by Arbitrum. In theory, a spamming attacker could halt Rollup's progress by launching a series of successive challenges, each of which would take a considerable amount of time to resolve. In fact, this problem plagued the last iteration of Arbitrum.

However, an updated version of Arbitrum provides an elegant solution to this problem, called Pipelining. The "pipelining" model allows the network validator to continue processing transactions for final approval even if the previously processed transaction is still in dispute. This creates a "pipeline" of transactions that have recently been processed but have not yet been finalized, rather than a bottleneck platform that prevents orders from processing transactions and network participants from submitting challenges.

The pipelined model works because anyone monitoring the network can know whether a dispute is valid or invalid before the dispute resolution process is complete. Essentially, the verifier can proceed with the default that the disputed transaction has been completed, and continue to build the rollup chain (that is, process the transaction) based on the correct result or "branch." This process, as shown in the following figure [2], reduces the potential for any spamming attacks. [3]

Source: Offchain Labs Development Center

conclusion

In addition to the design of the dispute resolution process, there are other significant differences between Arbitrum and Optimism, in particular

• 他们的代码库架构，以及

• 他们对矿工可提取价值 (MEV) 的方法

First, a brief summary of these differences: While the Optimistic codebase is relatively simple, the Arbitrum codebase is more complex and ambitious; Optimism has stated in the past that it prefers the MEV auction method, while Arbitrum plans to provide the FFS, Fair Sequencing Service. Of course, these two comparative views deserve to be discussed in detail in separate articles. MEV, in particular, is a matter of philosophical debate between the two projects -- although both projects are expected to use a Trusted Sequencer Model for simplicity, at least in the early days after launch.

Finally, taking a step back from the protocol nuances (important though they are), there are some "soft" elements that separate these two heavyweight projects: guiding strategy, motivating design, community spirit, and so on. In fact, Optimistic Rollups must have an ecology of their own if they are to be successful in the long term, and not just an add-on to Ethereum. So expansion is more of a multifront war than an arms race. There may be only one winner in this war, or there may be multiple winners. It may last for years, or it may end very quickly.

But it will certainly have a major impact on the future of the crypto world.

This article does not make any investment recommendations.

[1] If the transaction uses contracts that have never been part of the proof of fraud before, then these contracts all need to be redeployed to Ethereum, which consumes a lot of gas.

[2] "DA" stands for "disputed assertion," which is a technical term for a transaction that has been processed but may be subject to challenge.

[3] You should also listen to Ed Felten's talk on pipeline processes at ETHDenver last year.