原文标题：《 Building Back Better with OP Stack 》
Original article by Donovan Choy
The DeFi way
There is a crypto bull market in 2021. Ethereum became crowded, with Gas bills sometimes running into the hundreds of dollars.
So what was the solution? Deploy the new faster chain! "Ethereum killers" like Solana, Binance Smart Chain, and Avalanche all emerged at that time. They've all done liquidity mining, and they've all pulled money out of Ethereum.
People have abandoned Ethereum despite supporting it in the past.
The Alt-L1 chain temporarily addresses the market's need for scalability. But the emergence of so many chains with different standards creates the need for bridging across chains. This, of course, brings trouble, not just inconvenience, but danger. Just as nbsp; As Rekt lists, cross-chain bridge hacking incidents are frequent.
Ethereum tries to answer the scalability question in a different way with rollups. Allocate transactions to a separate (rollup) chain for processing and inherit the decentralization and strong security of the mainnet.
It has been working. As of October, L2s had surpassed the Ethereum mainnet in transaction execution.
There is still a problem, however. Each new chain or rollup is itself a single chain, with its own independent language and design structure. The result of extending this over time will be an asynchronous Web3 ecosystem. This is why we need a protocol like Hop to bridge between rollup chains.
If Web3 is to be adopted by the mainstream, we need to say goodbye to this constant network switching. People don't want to deal with fragmented, asynchronous chains. It's like having to exchange currency every time you want to make an interbank transfer.
If cross-chain Bridges are disposable Band-aids, modular rollup is paving the way for more comprehensive solutions.
So how to proceed? The developers behind Optimism have a cure: the OP Stack.
Today, most chains are pursuing their own modular strategies. Optimism, Arbitrum, and Starknet differ in their execution layers, but they all share the same layers of clearing, consensus, and data availability by outsourcing them to the Ethereum mainnet.
Metis and Celestia choose their own data availability layer, while still using Ethereum for their settlement and consensus layer. Starkex-based validium chains, such as Immutable X or rhino.fi, do something similar by running their own relatively centralized data availability boards - pre-selected sets of nodes that host transaction data.
Each chain runs its own separate, single policy, with differences in execution, billing, and design of the data availability layer. What if these chains shared a standardized open source code library instead of today's isolated chain /rollup system?
This is where Optimism OP Stack comes in: a modular set of basic Lego building blocks for building more expressive and precise rollup chains on Ethereum that are not possible on today's monomer L2.
The OP Stack is a standardized set of open source modules that can be assembled into a custom chain -- what Optimism calls an "OP chain" -- to serve any particular blockchain use case.
Let's take this definition apart:
A module is a bit of data that any developer can plug into the OP Stack to create an L2, L3, or L4. "Standardization" means that there is consensus on the standard of a module and that it can be implemented by all. Open source means it is free for anyone to iterate and request.
With OP Stack, you are not tied to a particular proof system or technology. Developers have the ability to switch modules between different execution, consensus, settlement, and data availability layers in a chain, just as they switch apis.
dYdX chose to leave Ethereum for the Cosmos application chain because they wanted greater modularity in the consensus layer on their chain. OP Stack solves this problem.
OP Stack is designed to allow for code forks in a much easier way than current efforts, as developers can easily abstract away the various components of the blockchain and modify it by inserting different modules.
If an Optimistic rollup wants to transform himself into a ZK rollup. No problem! Just change its fraud proof module to the validity proof module of the settlement layer.
If a chain wants to use Celestia for its data availability layer. No problem! Just swap Ethereum for Celestia.
Do you want to swap the executive EVM for another virtual machine, such as FuelVM? This is difficult on a running chain, but it's a technical possibility for OP Stack.
Maybe you want to run Minecraft as an L2 rollup, but the on-chain game is too computable on the mainnet? In fact, someone has already done it, namely OPCraft, the team behind Lattice by introducing its own execution module into the execution slot of the L2 rollup, and then modifying a plasma in the consensus layer to increase scalability.
Thus, OPCraft acts as its own L2 rollup (OP chain!) Existing on Ethereum, every action in the game is executed as an on-chain transaction and scrolls down to the Ethereum mainnet. It's Minecraft on the EVM-compatible blockchain, and it's extensible! Just like other blockchains, developers can access it through nodes and deploy smart contracts on it (see this Metatarsal post for more on thisarticle.)
Lattice does this by utilizing the optimistic Bedrock rollup architecture. Bedrock, the first implementation of the OP Stack, is a collection of modules used with Optimism. Bedrock uses an Ethereum virtual machine as an execution layer, making it equivalent to an EVM, and Cannon as its interactive proof of failure system on the settlement layer.
More crazy forks are coming. 0xPARC sets up a Game Boy rollup by replacing Bedrock's execution engine with a Game Boy emulator.
This is all done on the chain.
Today, the main problem with modular blockchains is increased fragmentation, as developers design choices and tradeoffs. This fragmentation problem is similar to Web2's walled garden, only here it is accidental.
OP Stack solves this problem of gradual fragmentation by starting from the idea of an open garden and building in Web3. As long as OP chains voluntarily choose to enter the same shared sequencer set (the only entity on each OP chain that generates blocks), all OP chains can enjoy atomic cross-chain combinations.
If OP chains may not want to run their own sequencer, then they can pay a fee to use the Optimism of their trusted shared sequencer. This opens up an alternative to Optimism as a source of Optimism, not just the dapp that is currently part of the optimism chain.
Ultimately, any user on Ethereum can send transactions to each other from any end of the ecosystem. No more network switching or bridging required!
This vision has given birth to the emerging structure of the Optimism "Superchain", where hundreds/thousands of OP chains will be fully interoperable on Optimism and connected by the same technical structure.
Launching rollup will be no more difficult than launching ERC20 tokens, and the pace of Web3 experimentation and innovation will be further accelerated.
Of course, it's not just about interoperability.
As the OP Stack shared module makes configuration more flexible, developers are reclaiming reusable code that was used by previous developers, making the code more powerful and more resistant to hackers and bugs.
For example, when the Lattice team built OPCraft, they designed each block with a much higher gas limit than Optimism of its own chain. In this different configuration, they found some bugs that were not obvious before.
All in all, OP Stack is an ode to Ethereum's basic vision of scaling through modularity.
Optimism is building on the optimism of establishing a fully open source rollup chain ecosystem, and OP Stack is the foundation on which this vision is built. But it's not just rollup, the OP Stack can be leveraged to build governance and identity modules, giving developers the ability to easily design their chains from scratch.
The OP chain is not a Cosmos application chain. They're not single-stranded. It is a super chain of the same technical standards.
When all this is done, the rollup ecosystem will blossom on Ethereum.
Disclaimer: OP Stack is still in early stages of development. The Optimism team is working hard to document and clean up the API to achieve this goal. While the codebase can be easily forked by motivated teams like Lattice, it will take some time before it is ready for use. Finally, a special thanks to the Optimism team for their comments on this article.