Source | polynya

Throughout the first ten years of the blockchain industry, only a single type of blockchain existed. Experiments like early plasma, multi-chain, and sharding tried to break through this situation, but it was not until the emergence of rollups, validiums, and data availability chains that heralded the end of the single blockchain era. However, we are still limited to a "single" perspective and use limited vocabularies such as L1 and L2 to describe revolutionary new design areas. Here's a thought experiment from a few months ago, using more descriptive nomenclature.

I think that if we want to understand blockchain modularity and blockchain Lego, we need to change our perspective. I am not sure which description is more appropriate, "modular" or "Lego". What do readers think? Do you have a better meme?

So first, what is a monolithic blockchain? In simple terms, a blockchain system has three tasks: execution, security, and data availability. For a long time, the blockchain system has been required to take on these three responsibilities alone, which has led to serious inefficiencies, which are also reflected in the blockchain's impossible triangle dilemma. Bitcoin and Ethereum chose high security and decentralization at the expense of scalability; while other chains made different tradeoffs.

implement

implement

This is where users interact and where all transactions happen. For end users, this layer has no difference in experience from a single-type blockchain and can be directly compared.

The execution-specific layer focuses on processing transactions as quickly as possible, while "outsourcing" the challenging work of security and data availability to other projects.

Rollups are the primary execution layer, but we also have validiums and volitions. Currently, Arbitrum One has a clear first-mover advantage in the market, followed by Optimistic Ethereum. But both A1 and OE are in their early stages and lack basic calldata compression optimizations such as signature aggregation.

StarkNet has now been on the public testnet for 3 months and is getting closer to MVP. I believe the last major hurdle is broad compatibility with components like web3 wallets, account contracts, etc. The predecessor of StarkNet, StarkEx, has implemented calldata compression technology, and signature aggregation is the default function of zkRs, so the transaction fee will be much lower than the current ORs, such as the average transaction fee of dYdX transactions < 0.2 US dollars. Even if Arbitrum One were able to implement these optimizations in time, zkR would fundamentally achieve calldata compression better than OR. StarkWare believes that StarkNet v1 will be released on the mainnet by the end of the year with a version compatible with the EVM through the Warp translator, but it is conservatively said that it is likely to be released by early 2022 at the latest. Another advantage of StarkNet is that it is actually a volition, not a rollup, but we are still waiting for more details on this.

zkSync 2.0 is another promising EVM-compatible zkR. Oh, it's not actually rollup, but a volition scheme similar to StarkNet. We do have more details about the architecture of zkSync 2.0 though. Arbitrum One, as a rollup itself performs all operations, but relies on Ethereum for both security and data availability. However, Ethereum is expensive as a data availability layer. So what volition does is let users choose between data availability on ethereum (rollup mode) and data availability on a different chain (validium mode). zkSync 2.0 will have its own data availability chain, zkPorter. Rollup mode is still the safest option, and zkPorter will offer very low fees (imagine ~$0.0X) while still being more secure than sidechains and other monolithic blockchains. We can already observe from Immutable X. My expectation is that zkSync 2.0 will release a public testnet this month and launch the mainnet in early 2022. Note, however, that there is always a certain delay with cutting-edge technology.

safety

safety

To be precise, I understand "security" to mean "consensus", but I think the word "security" is better to avoid confusion between the execution layer and the data availability layer, because the latter is not necessarily without a consensus mechanism.

Of the three, "Security" is by far the hardest layer. Currently, there are only two sufficiently secure, decentralized (or attempting to be) solutions: Bitcoin and Ethereum. Most other chains didn’t see the blockchain LEGO coming and made huge sacrifices in security and/or decentralization to achieve higher scalability.

data availability

data availability

Long term, Ethereum also has the best roadmap for data availability, both in terms of KZG promise technology and data availability sampling, and purely utilizing its industry-leading security to deploy massive data shard chains.

But ethereum's data availability layer may have to wait about 18 months. In the short term, validiums and volitions schemes could build on Ethereum's security while submitting transaction data (in compressed form) to a separate data availability layer. We have data availability chains like Polygon Avail, Celestia, and zkPorter, and committees like StarkEx's DAC, who will fill the void and have the opportunity to build network effects. It should be noted that some of these chains also focus on their security, but as mentioned above, I don't think these chains will compete with Ethereum in this regard.

As for candidates outside the Ethereum ecosystem, we may also see monolithic blockchains such as Tezos and NEAR providing sharded data availability before Ethereum. Although these chains are significantly inferior to Ethereum in terms of security and decentralization, they can serve as data availability chains.

epilogue

epilogue

We are entering a bold new era of blockchain LEGO, bringing order-of-magnitude efficiency gains to the industry. I hope this article outlines the future competitive landscape. Monolithic blockchains are all but obsolete, they need to focus on execution, security, or data availability, and it is impossible to compete if they still try to satisfy all of these characteristics. Projects that have already chosen tracks, such as those listed above, will be big winners in the coming years and deserve attention and support. I expect that in the coming months and years, the exponential efficiency gains from the modular model compared to the monolithic model will be obvious to all, and people will rush to enter this field, especially is the execution chain field.

Original link: https://polynya.medium.com/the-lay-of-the-modular-blockchain-land-d937f7df4884