Extended thoughts on the DA characteristics of Plasma

Author: Haotian, Crypto Observer Source: X (formerly Twitter) @tmel0211

In short: Plasma’s Data Availability (DA) processing solution is extremely immature, and adopting Zero-Knowledge (ZK) technology is a last resort. There are inherent “flaws” in those attempts to bypass ZK and implement clever DA schemes in the name of Plasma alone. Here’s a detailed addition to this view:

1. DA validity of Plasma: Since Plasma only publishes Layer 2 block headers/Merkle Root on Layer 1, most of the DA data is located off-chain in Layer 2. This means that the Layer 2 Sequencer can withhold the data and refuse the validators of the mainnet to verify its full DA data, resulting in the loss of validity verification of the exit game withdrawal request initiated by the user.

2. Mainnet Validators that have lost their validation: Once the Mainnet Validators have lost their validation, the user’s exit game withdrawal becomes complicated. Layer2 Sequencer can insert some counterfeit transactions, interfering with the validator’s ability to validate, and even the existence of a “withdrawal priority” mechanism cannot stop Sequencer from doing evil.

3. Limitations of UTXO’s account model: Although the UTXO model can sort transactions and provide exit game enforcement reference for mainnet validators, it cannot fine-grained management of the “ownerless state” of smart contracts. This also makes it difficult for the Plasma-based smart contract Layer2 solution to truly solve the problem of exit game evil.

4. Redstone’s DA solution: Redstone files the datahash of the off-chain data before the block header is uploaded to the mainnet, and users can launch a challenge to verify the original data. This barely solves the problem of data retention, but there is still the issue of the authenticity of the data itself.

5. The necessity of ZK technology: Only through ZK technology can we truly solve the problem of Sequencer’s potential evil. Because if the state data submitted to the mainnet does not have the corresponding DA data, such a sroot will be rejected as invalid, which fundamentally eliminates the data retention problem in Sequencer.

To sum up, the only direction that can lead Plasma out of the innate quagmire of the DA problem is the ZK technology. Redstone’s Plasma-inspired DA solution solves the problem of “data retention” by consensus, but it does not really solve the problem of Sequencer’s centralization.