Original author: Haotian (X: @tmel0211)
Related Reading:
《V God’s long article: EVM Validiums’ exit from the game and the return of Plasma》
After reading @VitalikButerin’s new work on the return of Plasma, I was deeply attracted by Plasma’s UTXO-based ledger “exit game” mechanism, and Vitalik seems to be interested in guiding the market to explore in the direction of ZK+Plasma to avoid the market staying in the Rollup stage. Next, I will give you a detailed popular science explanation:
Key points: 1) Why is Plasma chain suitable for payment scenarios? 2) How does the Plasma payment scenario work? 3) Why is the exit game mechanic important to Plasma? 4) Why is it difficult for the Plasma chain to integrate the ownerless state of EVM? 5) What imagination space can ZK+Plasma release?
I am herePrevious article Interpretation of VitalikAs mentioned in the article, Ethereum’s layer 2 expansion solutions originally included: Plasma, Rollup, Validium, Parallel and other solutions.Vitaliks expected expansion direction should be balanced development, adapting to various application scenarios and diversifying layer 2 construction. However, the actual market structure is that the Rollup solution is the only one, and it is increasingly involution.
However, although the Rollup solution has high security, it relies too much on Data Availability, and pure Ethereum DA is limited by performance and cost issues. Therefore, the market now has a stuck Rollup camp but relies on third-party DA. Solution, which is deforming Rollup, is definitely not the status quo Vitalik wants to see.
Therefore, Vitaliks new article re-examined Plasma and introduced a ZK+Plasma expansion solution, which is obviously another layer 2 political flag-waving issue.
Why is Plasma currently limited to payment scenarios?
Plasma is equivalent to a side chain solution that regularly synchronizes Merkel status data to the main network. It is an expansion solution that relies on the main network for data and computing. In this case, Layer 2 can use a very centralized approach and design a very complex ledger model to efficiently expand, and it can also reuse the system capabilities of the main network vailidator.
Generally speaking,Plasma is used in payment scenarios to ensure that the ledger status is effectively tracked and recorded.why?
1) In payment scenarios, users only need to retain the balance status. If all off-chain status data must be retained in other scenarios, it will bring pressure on storage space;
2) Plasma security relies on the exit game mechanism. If the operator does evil, users can initiate challenges and submit their own assets. If the asset status is complex, it will be very troublesome;
3) At present, it is difficult for Plasma to be compatible with many ownerless states of EVM, and it is difficult for users to use Plasma’s Merkle state tree ledger to correspond to many non-transaction states, such as LP and CDP;
How does the Plasma payment scenario work?
As stated in Vitaliks article, in simple terms:
Plasma Cash can treat each Token as an NFT with a unique number. When a user makes a transfer, the operator will record an updated status on the Merkle leaf. Each user can save his or her global Merkle. Tree state, so that the ledger will not be messed up when traced back.
If it is a homogeneous Token, users may split and merge it multiple times when consuming it. For example, Xiao Wang first split 1 ETH into three parts, and then merged two of them. Each part is different. Regular, such as 0.001, 0.1, 0.3, etc. If split in large batches, it may lead to Merkel data redundancy, which may cause problems when initiating the exit game mechanism (higher challenge and verification costs), such as looking back nearly a week The ledger was found to be inconsistent. How to solve it?A UTXO-like ID can be matched for each asset split or merger status, so that no matter how the asset is split, the corresponding transaction leaves can be immediately located.
How to ensure a safe exit game mechanism?
Because Plasma does not have its own independent chain system like Rollups, it has to ensure that its side chain accounting is synchronized with the main network at all times. This allows it to not deliberately pursue decentralization, as long as it has an operation that can efficiently record accounts. Just the main body.
But here comes the question, what if the operator releases invalid blocks, records false accounts, and steals user assets?Users need to issue an exit game mechanism at any time to withdraw assets from layer 2 back to layer 1, similar to Rollups escape hatch safety mechanism.
How to do it? Users can post proof of their Merkle tree status, prove the asset transfer process, and initiate a 7-day challenge period.The main network verification node will detect whether the user is the final asset owner and whether the user has problems such as double spending of assets.(Because the main network node stores more Merkle tree states, it can check whether the users Proof is suspected of malicious exit).
By exiting the game, the layer 2 operators are restrained from doing evil, and the existence of the challenge period prevents users from maliciously exiting, which ensures the normal operation of the Plasma chain.
What are the difficulties in being compatible with EVM ownerless state?
As mentioned before,Plasma is currently more of a 2-layer solution designed for payment transaction scenarios. It is an accounting that can be compared to the UTXO model, and EVM itself is an account model.UTXO can record every balance status refresh, but many Owner-less scenarios in the EVM state machine are difficult to implement using Plasmas solution.
For example, USDC deposited in the Uniswap pool is the same as the assets placed in MakerDAOs CDP. It is difficult for users to prove which assets belong to them. In this way, once the operator goes down and the contract is locked, the user cannot function normally. exit the game.
Because there is no way to prove that the user has money in the contract. Affected by the characteristics of Plasma data, the main network can only monitor the balance of the contract. If the operator issues an additional amount of money to the contract, how can the user prove which amount is their own and which amount was maliciously issued?
Whats more, if a layer 2 side chain can only realize payment transfer transactions, how to build applications and ecosystems? Obviously, this will greatly limit the usage scenarios of Plasma.
The imagination space released by ZK+Plasma
If the bottom layer of Plasma is completely ZK-ified, user operations will exist in the form of zk-SNARK proofs, which can release many EVM state machine scenarios:
If a user deposits an asset into a Plasma contract, a zkSNARKs certificate can be constructed.This proof can initiate an exit game to the main network, so that even if the pool is frozen due to security threats, users can withdraw their legal assets;
Or in transaction scenarios involving privacy DEX, users can use zkSNARKs to prove ownership of an asset without exposing their privacy; and Plasmas smart contracts can also use zkSNARKs to prove the correctness of their state upgrades when performing complex upgrades. Without exposing details, it can also make it more difficult to do evil in a contract;
above.
Overall, Vitaliks entire text clearly describes the current status and existing problems of Plasma, including the possibility of future ZKization.In my opinion, Plasma is not new. It has even found its place in the payment scene over the past many years. At this moment, Vitalik throws out the new direction of ZK+Plasma, which is a direction guide and a political flag-waving.As for whether the market will follow Vitaliks wishes, I personally am not very optimistic:
1) Rollup is the market’s optimal solution based on investment costs, development difficulty, ecological compatibility, etc. ZK+Plasma is certainly a more advanced ZK-Rollup, but the current development of ZK-Rollup is not optimistic. Jumping to Plasma is too hasty;
2) Validium, as an independent chain expansion solution, is relatively more advanced in ZK’s application, but it completely relies on off-chain DA. In comparison,ZK+Plasma seems to be more sticky with the Ethereum main network, but Vitalik’s call is understandable. ZK’s mature developers may not have enough motivation to abandon Validium and engage in ZK+Plasma.
Note: Please be sure to read Vitaliks article carefully before understanding this article.
