Pushing Past Limits
Lattice was founded with one key mission: build software that accelerates timelines, enabling developers to ship ambitious onchain worlds and applications. Unsurprisingly, these applications often push Ethereum and the EVM to its limits. When we built MUD, our framework for ambitious onchain applications, some of our earliest in-house experimentations were on a local chain with a 1 billion gas limit, to really interrogate just how much state and throughput a MUD application could handle. We choose to live in the future, because we believe that the EVM one day will be able to handle far more data than it does today.
As the MUD framework has reached more users, we’ve seen more applications launched, and the rise of new kinds of games and nascent autonomous worlds that would have been impossible to build before. The applications we’ve seen built with MUD: Sky Strife, OPCraft, Words3, Primodium, draw.tech, Gaul, THIS CURSED MACHINE, just to name a few, have made us confident that MUD unlocks onchain applications that would not have been feasible otherwise. But most MUD applications today are still on testnets, and we need a path to bring these apps to a chain that can support onchain worlds.
We first began to suspect that L2s might struggle to support large-scale MUD worlds when we deployed OPCraft on an Optimism Bedrock testnet last year. While OPCraft was a great illustration of the power of building with MUD, it also inspired some trepidation in us: it would’ve been wildly expensive to place a block in the game if we had been posting transactions to Ethereum mainnet from our OP Chain, and the game was saturating blockspace during periods of intense activity.
This led to a realization: if current L2 offerings weren’t sufficient for the kinds of applications we know are possible with MUD, we would have to build one.
When OPCraft was still called “MUDCraft”, we began researching the optimal way to scale transactions in MUD worlds. Our research brought us deep into the early history of Ethereum scaling: Plasma. While Plasma never got serious traction, one aspect of the design caught our attention: instead of posting all transaction data to L1, Plasma used data commitment hashes combined with offchain data availability and onchain challenges. The things that prevented Plasma from being a practical scaling solution (primarily the lack of smart contract support) could be alleviated if we built a Plasma-like alt-DA protocol on top of a rollup like the OP Stack.
So we decided to do that. Earlier this year, we began iterating on a Plasma-inspired alt-DA design, built on top of the OP Stack. Within a few months, we had multiple conversations with Karl Floersch — co-founder of Optimism and now CEO of OP Labs — brainstorming a way to build L2s with cheaper L1 security fees using old ideas from Plasma. Over two weeks of an intense coding sprint in Paris, the first proof of concept of Redstone was born.
Today, we’re announcing Redstone, the first plasma-implementation for the OP Stack. Redstone is a super cost-effective chain for onchain games, worlds and other ambitious applications. It operates like a traditional optimistic rollup, except instead of posting the input state to L1, we post a data commitment hash. The input state corresponding to the input commitment is stored offchain by a Data Availability Provider. To ensure the input state corresponding to the input commitment is available, there is a Data Availability Challenge contract on L1, which allows anyone to submit a challenge to a data commitment in case the provider misbehaves.
The challenge system we’ve designed for Redstone is permissionless, meaning that any commitment can be challenged by any user: any honest actor on the internet can enforce the availability of the chain. Thus, Redstone does not rely on a permissioned set of data validators, or an alternative consensus mechanism outside Ethereum. We use Ethereum itself to force data to be available.
Onchain worlds need a sufficiently secure chain, opening the door for non-Ethereum data availability. Just like MUD, we want Redstone to use primitives from Ethereum and to be permisionless. Redstone leverages Plasma-like data availability, allowing any honest actor on the internet to make sure the chain is available, without requiring new forms of consensus outside of Ethereum.
Whereas regular optimistic rollups post the entire input data on L1, spanning 120 kilobytes of data per batch, the first version of Redstone hashes each batch, bringing down the data posted to Ethereum mainnet to 32 bytes. As the amount of input data is responsible for the bulk of a rollup’s L1 security cost, Redstone is able to bring the L1 security cost down by more than 90%. In future upgrades of Redstone, we can reduce this cost even further through merklelized commitments, which allow you to hash multiple batches of 120 kilobytes into 32 bytes, rather than just one batch.
Redstone was designed by the Lattice team, with collaboration from Optimism. With Redstone, we are joining as a contributor to the Superchain, and the OP Stack as core developers. We chose to build on the OP Stack because we share the team’s commitment to fully open-source software, as well as the same principles of designing common standards to allow independent teams to build modular integrations shared by everyone (instead of isolated, custom solutions). Other tenets of the Superchain, like enabling horizontal scalability by spreading state and users across multiple chains, is something we can leverage for Redstone, to allow players to inhabit apps and worlds far beyond current limits today, with underlying shared security and fast messaging across chains.
By joining the Superchain we hope to accelerate the adoption of not only Redstone, but MUD. We’ve designed Redstone to be highly compatible with MUD, with optimized RPCs and indexers that will ease the developer experience for builders creating ambitious applications and worlds.
Today, the Redstone testnet is launching publicly. In addition to Sky Strife, our in-house onchain RTS game, other MUD builders soon launching games on the testnet include Small Brain Games, Gaul, and Primodium. Redstone mainnet will launch later next year.
Redstone brings us closer to our ambitions of pushing the limits of what is possible to build on the EVM. Coupled with applications built with MUD, we believe Redstone brings us closer to living in the future. If you would like to build on Redstone, please reach out at email@example.com.