The Future is Modular

Until recently, blockchain infrastructure was mostly monolithic. In other words, all components of infrastructure, including consensus mechanisms, data storage, and smart contract logic, were built as a singular system.

This lack of modularity has been a major drawback for Web3’s prospects, as builders have had to create their applications around the inflexible requirements and constraints of existing chains, networks, and nodes.

Modularity rose in prominence as developers wanted more flexibility and variety; they wanted systems’ components to be able to be separated and recombined.

Picking and choosing core components like transaction execution, transaction ordering, and data availability became important for builders who wanted to customize the stack for their use case.

As Web3 infrastructure matured, this evolution toward modularity became evident in a few areas, from the Ethereum roadmap and proliferation of Layer 2s (L2s )and Layer 3s (L3s), to application-specific chains (app-chains), and more.

Since the early days of blockchain development, Ethereum has grown into the “Mother Tree” of the Web3 forest. It’s evolved to become highly robust, yet less and less flexible to any major changes.

Just as the roots of a tree anchor it to the ground, transport nutrients, and provide stability, Ethereum validators secure the network, validate transactions, and maintain the integrity of the blockchain. In other words, the root system (validators) is the lifeblood of the tree’s survival.

Over the years, the larger and more distributed its root system became, the more robust and anchored the tree became. Everyone wanted to grow its leaves (applications) on top of the Ethereum tree. The more leaves the tree grew, the more congested the tree's transportation system (i.e., network) became - giving rise to the L2s.

L2s made Ethereum more scalable. In the tree analogy, we can think of L2s as the core branches sprouting from the Ethereum trunk. Fundamentally, they couldn’t exist without the trunk (L1). They rely on the tree’s robust root system (validator set) for transaction settlement, but they’re a whole lot more flexible than the trunk.

For example, as an application developer, you can pick whether you want an L2 with the fastest throughput of 65,000 TPS (e.g., Polygon) or the lowest computational costs (e.g., Arbitrum).

This was one of the first steps toward “modularity.”

The application developers wanted even more architectural flexibility - the ability to customize different elements of their chain - which is what drove the need for modularized L2-specific app chains. We can call them Layer 3s (L3s).

In our tree analogy, L3s represent the “sub-branches,” or twigs, growing from the larger branches (L2) of the Ethereum tree (L1). Similar to how twigs are the most flexible parts of the tree, L2-specific app chains give the developer the most flexibility regarding the architectural choices they can make while building an application.

For example, an app-chain developer might choose to modify transaction ordering by batching certain transactions together to give users fairer prices (and thus reduce MEV).

Optimism’s OP stack and Superchain are the most mature examples of the “L3” product. The OP Stack has already gained meaningful traction - it’s being used to launch Coinbase’s Base chain, Zora Network, a new Binance chain, a public goods network by Gitcoin, and now Celo.

A competing L2, Polygon, has also announced a similar L2 proliferation strategy for Polygon 2.0, seeking to be a network of ZK-powered L2 chains, unified via a cross-chain coordination protocol.

Last but not least, the current kingmaker of the L2 chains, Arbitrum, has taken a different approach by adding a layer. They’re enabling builders to launch new L3s known as Arbitrum Orbit chains, which settle first to Arbitrum, then eventually to Ethereum.

In a preemptive effort to anticipate demand for the rise of app-specific roll-ups, we’re even seeing the rise of “roll-up as a service” providers, like Caldera, Eclipse, Cartesi, and Gelato.

All these efforts are a clear sign that the future of Web3 development is modular, not monolithic.

Clearly, the Ethereum “tree” is now rapidly growing in size, biodiversity, and flexibility - with more branches, twigs, and leaves of various colors and sizes.

But there’s more to the Web3 forest than one beautifully biodiverse Ethereum tree.

Just as different types of trees and plants contribute to the diverse ecosystem of a forest, there’s a great diversity and efficiency growing within the blockchain ecosystem.

This biodiversity of the Web3 ecosystem now includes hundreds of L1 blockchains - including 450+ standalone L1s (i.e, Avalanche, Solana, etc), as well as 60+ application-specific chains powered by Cosmos-SDK (i.e., Osmosis).

The latter trend cannot be underestimated, with large incumbent applications, like DyDx, or institutions, like Circle, launching their own app-specific infrastructure to access a feature set that general-purpose infrastructure cannot accommodate.

This further underlines the move towards modularization.

Just as the trees, plants, and other organisms in a forest are interconnected through various relationships, like symbiosis and competition, modular components in the blockchain ecosystem are growing to interact with each other - making this a highly “composable” forest.

It’s now growing into a “LEGO forest,” wherein key components of the blockchain stack are also breaking off into a “plug and play” model.

One example of this move towards the “plug and play” model is evident in the rise of the data availability (DA) layer solutions, like Celestia, Eigenlayer (Bloccelerate portfolio), and others. Just like one LEGO block can be stacked on top of another, these chains will be leveraged for the “data availability” functionality for other execution layers (e.g., roll-ups).

For more details on data availability, __this __is a great blog to read.

This analogy extends further with the new “restaking” primitive - this was a particularly popular topic at ETH CC. Restaking enables the validators of Ethereum to extend their service to other applications and app chains - as if the roots of the Mother Tree are now able to support the growth of other trees and plants in the forest.

Eigenlayer is also leading the restaking revolution, allowing applications to leverage the economic security of Ethereum instead of having to bootstrap their own validator networks (explained in an analogy by Sam Yilmaz). There are a lot of exciting new developments, and modularity will enable the next wave of exciting blockchain applications.

Lastly, the Web3 middleware infrastructure, like wallets, is also becoming increasingly modular and easy to integrate with (e.g. Metamask Snaps allowing any chain to build its own Metamask integration).

What are the key takeaways from this activity?

It’s becoming increasingly clear that the Web3 ecosystem is growing into a true Amazon basin rainforest, where different species, organisms, and plants interlink in a complex and interconnected web of life.

As this trend continues, there’s going to be a growing need for the various species - or blockchains and/or app-chains - to co-exist. This means cross-chain interoperability will only become more important.

As the Web3 ecosystem is growing to be more modular, the builders will have increasing options for customization and modularity, but the jury is still out on whether or not most of the plants, or even the branches (L2s) and twigs (L3s) of the Ethereum tree will survive.

Winters are getting harsher in the Web3 forest, after all. There’s a possibility only a handful of them will make it, while the rest will die from malnutrition.