I see that the quick confirmation function of Uniswap deployed on OptimismPBC has aroused many people’s interest. But how is this done? Can users rest assured to use it? Doesn’t relying on a sequencer to provide confirmations threaten decentralization? Let me answer you one by one.
First, and most importantly, the sequencer is a privileged participant in many rollup systems (@optimismPBC, @arbitrum, @StarkWareLtd, @zksync). They receive transactions from users, sort them and submit them to Layer1 in batches.
The sequencer exists mainly because a single coordinator is simple and efficient. At this stage, each rollup system usually has a sequencer, which is run by the system creator.
The sequencer is responsible for sorting transactions. Therefore, after receiving the user’s transaction, the sequencer can immediately mine it out and return a confirmation to the user. This greatly improves the user experience.
If you are worried that the sequencer will grab MEV, you are right, but I will discuss this issue separately (for the final analysis).
If the sequencer is faithful to its duties, everything is fine. But what should we do if the sequencer does evil, deceives users and tries to disrupt the network? Let’s explore this issue in depth.
The most important question is: can the sequencer steal the user’s funds? can not. The validity of the state transition is guaranteed by the rollup architecture (Optimistic Rollup relies on fraud proof, zk-Rollup relies on validity proof).
Can the sequencer review user transactions? Yes, it does. The sequencer is usually a JSON RPC node. Similar to Infura, the sequencer can even falsely report network status or censor user transactions.
Fortunately, censorship is not a big issue, because all rollup systems can publish Layer 2 transactions through the uncensorable Layer 1. The agreement will force the sequencer to pack user transactions into a rollup within a few minutes.
If the sequencer falsely reports the status, the user needs to run the node by himself, and recreate the rollup state based on the batch transactions posted to Layer 1. This may sound bad, but it is the same as the situation on Layer1.
Finally, can the sequencer falsely claim that the transaction has been confirmed instantly? Can. As mentioned above, the sequencer can falsely report the current network status and whether user transactions are packaged.
For example, the sequencer can lie to the user that the transaction was successful, but was actually cancelled. Only after the user recreates the rollup state based on Layer1 will he find himself cheated.
Only when it is published on Layer 1, the rollup transaction is considered finalized. This is why Rollup’s Web 3.0 library generally allows developers to easily build user interfaces to inform users of the progress of Rollup transactions.
A solution that may be adopted in the future is to have the sequencer sign and confirm when the user transaction is received. If the transaction is not packaged into the rollup, the user can punish the sequencer. This can be automated through services such as watchtowers.
This is what really excites me-sequencer technology is still in its early stages of development. In the future, we will see more complex designs to solve many of the problems I mentioned.
We can also run a license-free PoS network composed of sequencers instead of a single license-type sequencer. Each batch of transactions is packaged on Layer 1 by a sequencer randomly selected in the network. This will greatly enhance adversity and censorship resistance.
Of course, every sequencer needs to provide a deposit, and if it does evil, it will be fined.
Other projects, such as Arbitrum, are experimenting with a fair protocol to find the correct order of transactions. https://eprint.iacr.org/2020/269.pdf
Of course, it is also possible not to fight against MEV, but to embrace MEV: Participants obtain the right to run the sequencer for a period of time through bidding (but there are some problems with this idea).
If you want to know more about the design space of rollup, you can read the section “Who can submit bulk transactions” in this article by Vitalik ( Chinese translation ).
In short, the IMO sequencer has achieved a good balance between decentralization and speed. What we currently see is only the smallest viable product of the sequencer, and many smart people are working hard to improve it. The future will be bright!
Original link: https://twitter.com/krzKaczor/status/1415326134552641536
Author: Kris Kaczor Translation & Proofreading: Min Min & A Jian
Posted by:CoinYuppie，Reprinted with attribution to:https://coinyuppie.com/popular-science-sequencer-in-rollup/
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.