The biggest difference between Optimism and Arbitrum lies in the implementation of the fraud proof mechanism. Optimism uses a single round of fraud proof, and Arbitrum uses multiple rounds of fraud proof.
Recommended reading: ” Selected Links | Reading Arbitrum, a Popular Ethereum Expansion Player “
Original title: “Science | OptimismPBC vs Arbitrum”
Written by: Kris Kaczor
Translation and proofreading: Min Min and A Jian
Let’s first look at the common ground between Optimism and Arbitrum:
All belong to Rollup, which is a genuine Layer 2 solution, which stores all transaction data on Layer 1.
All are “optimistic”, that is, using fraud proof
Both use sequencers to achieve real-time “determinism”
All have general cross-chain messaging functions, which can be used to build advanced token bridges, such as MakerDAO’s fast withdrawal bridge
Next comes the main event-a different point . The biggest difference between Optimism and Arbitrum lies in the implementation of the fraud proof mechanism, that is, after a transaction is executed, how to resolve the dispute if the status is caused?
Optimism uses a single round of fraud proof. In other words, Layer 1 will execute the entire Layer 2 transaction on the chain to verify the state root. This makes fraud proof instantaneous, which is an advantage.
However, there are also some problems:
You need to supervise transaction execution, and for this you need to use OVM (that is, EVM rewritten to avoid side effects)
The gas usage of Layer 2 transactions must not exceed the gas limit of the Layer 1 block
Every time a transaction is executed, you need to generate a state root on the chain-leading to increased costs:(
There are security risks
Arbitrum uses multiple rounds of fraud proofs. Simply put, it is through binary search to find the first opcode of the block that caused the divergence. After finding it, just execute this opcode on the chain.
The advantages of multiple rounds of fraud proof are as follows:
The disadvantages are as follows:
Need EVM -> AVM translation (luckily it is automatic)
Very slow-in the worst case, it can take up to 2 weeks to complete the fraud proof. It’s actually a week.
Requires the original withdrawal to be online and cooperate
From another perspective, Optimism is containerized , while Arbitrum is virtualized .
Optimism’s method has a big disadvantage. Imagine if there is a hard fork that causes the Ethereum consensus rules to change and the opcode is deleted/repriced or modified. At this time, suddenly re-executing the past transaction on Layer 1 will result in a completely different final state. I don’t know how the Optimism team will solve this problem, but I think they will come up with a solution when the day comes. Arbitrum fully controls the AVM specification, so there is no need to worry about this problem.
Both of these projects try to get as close as possible to the Ethereum ecosystem, but there are some differences. In general, you can still use the EVM-related tools you know (solidity, hardhat, waffle, etc.). But it’s actually not that simple.
Optimism requires a special solidity compiler to generate OVM bytecode. Therefore, it is only compatible with Solidity and a specific version of Solidity. On the other hand, the Layer 2 node of Optimism is just a modified geth client. The purpose of the modification is to improve compatibility.
On the surface, Arbitrum is fully compatible with the EVM/JSON RPC specification, but its node is a custom implementation. Arbitrum supports fraud proof through automated EVM->AVM translation. Therefore, it supports any EVM language (vyper, YUL+, etc.).
Optimism uses wet and has a built-in wallet abstraction. Arbitrum comes with ETH support.
Arbitrum provides a unified permission-free bridge to achieve the bridge between any token and Layer 2 (it deploys a general ERC20 contract as the counterparty to Layer 2). Optimism prefers dedicated bridges, but it is also possible to build general-purpose bridges on Optimism. On this point, @dmihal knows more in detail 😉
The last difference is the online implementation. Arbitrum will launch the “Developer-oriented Mainnet” at the end of this month, while Optimism will have to wait until July.
If you want to learn more, I suggest you watch this friendly debate between @karl_dot_tech and @hkalodner hosted by @stonecoldpat0.
Personally, I support both of these projects. I can’t wait to see their mainnet go live. The entire Ethereum community urgently needs a suitable Layer 2 solution, not a scalable side chain.
If you want to understand the example of Optimism bridge, you can check it out . We are going to release an Arbitrum compatible version next week.
Disclaimer: As a blockchain information platform, the articles published on this site only represent the author’s personal views, and have nothing to do with the position of ChainNews. The information, opinions, etc. in the article are for reference only, and are not intended as or regarded as actual investment advice.
