Blockchain
Building blockchain-based applications often involves compromise at some point. Businesses must decide what they are willing to sacrifice – speed, cost, security, and so on – to achieve their specific priorities.
A service can choose to build on a layer-1 blockchain like Solana over Ethereum for speed and cost benefits, but lacks the network effect benefits of Ethereum. Alternatively, they can select a layer-2 rollup like Arbitrum or Optimism to stay in the Ethereum ecosystem but give up some of the benefits of having a direct presence on a layer-1 solution.
In a recent episode of the 0xResearch podcast, Eclipse Laboratories founder Neil Somani suggests it doesn’t have to be that way. Rollup solutions like Eclipse are designed to provide interoperability and customizability, so developers can vent too.
Somani acknowledges that monolithic layer-1s like Solana and Ethereum have an important advantage right now: “The only thing the monoliths have right now is network effects. They have a lot of liquidity on these chains. There are developers already writing for those platforms and who identify with the brand.”
But with the advent of rollup technology, the desire for another layer-1 to join the crowded space is diminishing, he says. “The downside is there aren’t many reasons why a new chain would choose to be a monolith over a rollup, assuming the rollup technology is fully built.”
Next-wave apps that “really push the boundaries” in the future, such as consumer-level apps and fully on-chain games, will require adjustments “at the execution layer,” Somani said.
“They are much better off with their own chain.”
The problem, he says, is that Tier 1 solutions are difficult to implement, very expensive, cumbersome, and require teams to ensure reliability. “So by implementing a rollup, they can avoid a lot of those pitfalls.”
A little history lesson
Somani explains a bit of the history of layer-1 development to provide context for the modular customization approach designed by Eclipse. “Ethereum gas costs have historically been very high, and there have been many attempts to fix that.”
Polygon and Binance Smart Chain, which were essentially EVM-compatible forks, emerged to solve the gas problem when it first reared its ugly head a few years ago. They “suddenly took off” purely because of timing with the gas crisis and the fact that no EVM-compatible Layer-1 alternative existed.
Fast tier 1 monoliths like Solana stepped in, along with the alternate vision of rollups like Arbitrum and Optimism. “This basically all solves the problem of limited block space, solving high gas costs.”
Other solutions, like the Cosmos-centric appchain vision, never really took off in quite the same way, says Somani. The value of Cosmos “isn’t really about lower gas costs” as there are plenty of tier-1 solutions, such as Solana, that already solve this problem. Instead, “it’s really more about the customizability that having an appchain gives you.”
Each approach has its own advantage over the others. “Like Cosmos, you get customizability. Solana, and these alternate L1s, you get low gas costs. And then the rollups get the benefit of sharing Ethereum security.”
Choose and take
Somani sees Eclipse combining these “core features” into a modular system that lets users “pick and choose and say, I want really low cost, but I also want to be part of the Ethereum ecosystem. Or maybe I want to be in the Cosmos ecosystem, and I’m fine with a more powerful execution node I just want as much throughput as possible.”
“So let’s people make those tradeoffs by breaking these different concepts or properties of a blockchain into these modules.”
Somani says the approach follows the same vision as Celestia, a modular solution for data availability layers. Prior to building Eclipse, Somani says Nick White, chief operating officer at Celestia Labs, reached out to him and explained what he saw as the future of blockchain technology.
White has broken down the complexity of blockchain into a few key features, says Somani. First, it needs to execute trades. Second, it must order transactions and make them available for verification over the network in processes called “consensus” and “data availability,” or “DA.” Finally, it must verify that state transitions have been performed correctly in a process called “settlement.”
“Eclipse just takes those three concepts and separates them very neatly,” says Somani.
“We let you deploy a custom execution chain, be it EVM, SVM, or any other virtual machine. You get to choose your consensus and DA tier.
“You can customize the DA layer,” he says, “and then you build those extra customizations on top of the execution layer to facilitate the kind of application you’re building.”
“Maybe you are a gamer and you need VR apps. You need verifiable random functions. Maybe you’re doing an NFT coin and don’t want to charge gas. Now you can make it gasless. You might want to do something special in the mempool, like redistribute MEV. These are all options you have when you have your own chain.”