Modularity Creates Adaptation: The Celestia Thesis

Modularity has been a popular concept for many years, specifically among designers and developers. This is because a properly designed modular system can not only generate endless variety due to its flexibility but also provide adaptation to satisfy the constantly changing and evolving demands of users.

If you look at the history of evolution, the term “fitness” does not mean the strongest. It means the one that has the best adaptation to its environment.

Layer 1's are the foundation of all crypto networks, and in order to reach meaningful scale, they need to change, evolve, and adapt to their environments over time. This is because the more people you have participating in a network, the more diversity of opinion and the stronger need for consensus.

Application Specific Chains

The introduction of Bitcoin in 2009 presented a novel way to coordinate large-scale economic activity, which gave birth to the idea of each blockchain having its own application. However, the way Bitcoin was designed made it extremely difficult to iterate and scale (no coordination between developers and validators). This was not an accident, it was designed to be extremely limited in its flexibility to deliver very robust security. With non-sovereign money, you don’t want the monetary policy to change; you want to have the properties of the system remain stable(and predictable) across time. Examining bitcoin’s original, application-specific design choice leads to insights on a different approach to scaling.

General Purpose Chains

The 2015 launch of Ethereum represented a pivotal shift in the way blockchains could be used and architected, by being general purpose. This system enabled developers to build new types of applications (dApps) and imagine all kinds of great experiences involving money, identity, and property that were not possible before (the technology didn’t exist to support them). But unlike Bitcoin, Ethereum was, and still is, constrained by design rather than by application. As a replicated system, it gains no resources as the number of users on the network increase, it loses resources. And although a general-purpose blockchain is strikingly different from the former, it is still subject to scaling limitations.

Back to Application Specific Chains

What the market is seeing now is a trend back towards application specific chains, but with a new model. Instead of having to endure the cost of bootstrapping a new POS network, developers can deploy their chains on pluggable consensus layers(drastically reducing the barrier of entry), similar to how virtual machines are deployed on AWS using physical servers.

All current Smart Contracts run on virtual machines (or an iteration of one), which means that all contracts on Ethereum share the same execution environment, which severely hinders their performance. For optimized performance, a native application should require its own chain.

I contend that Celestia is a solution to scaling the desirable properties of blockchains without sacrificing decentralization and security for speed.

Modular on Chain Data

The underlying problem that current Layer 1's face is that they are Monolithic, meaning they couple the execution rules for transactions with the consensus rules. In order for users to check if a block is valid they must download and process every transaction (this does not scale).

Celestia represents a paradigm shift in how we build, deploy, and use blockchains. It is a modular solution that separates and simplifies the core functions of what every blockchain needs to do in order to build applications.

1. Consensus + Data Availability
2. Execution

These two components can be specialized and layered on top of each other to achieve the same properties as (ETH,DOT, ADA, SOL,NEAR, ATOM) but with greater decentralization, security, and scale.

Checking that a transaction has been published i.e. “available” removes the need for the base layer to process and compute all the data. This gives developers the flexibility to define their own execution layers on top in the form of their own chains. You can think of a modular data layer as doing the minimal amount of processing to reach consensus, and allow applications to be built on top of it.

Celestia achieves this by simplifying block verification, to data availability verification.

Data Availability

Data availability can be understood intuitively; it simply means that data in a given block has been published to the network. It is a key concept in distributed systems because it is fundamental to the security of all chains. For example, when you want to check a chain’s validity, you must first identify all the transactions behind that state update. Data availability allows you to verify that transactions in a block are valid without downloading all of the data.

When current Layer 1’s try to scale, data availability becomes a problem. The reason you don’t hear about this is that it’s “solved” implicitly by all full nodes downloading the contents of every block before it is considered valid. This means that if a chain wants to have bigger blocks, it must have stronger processing nodes i.e. Solana. If the network requires more powerful nodes, anyone who wants to participate and verify the chain will need more resources, which tends to lead to centralization. Celestia sidesteps this problem by separating consensus and data availability from execution.

Data Availability Sampling

Data availability sampling is the technique implemented by Celestia that allows the production of bigger blocks (equivalent to increasing the gas limit for ETH) (more tx’s per second) without increasing the end user’s cost to validate that chain. This is because instead of full nodes checking if transactions are valid, they just need to check if they’re available. This incentivizes users with fewer resources to run nodes, and the more nodes you have in the network, the more you can secure a chain with larger block sizes.

In short, data availability sampling allows the size of blocks to increase in proportion to the number of nodes on the network.

Scaling

At a fundamental level, the purpose of scalability is to enable the application layer. This is what matters the most for a smart contract or Layer 1 chain. Everything built at the base layer is specifically for a better experience at the application layer.

Scalability is a function of cost (referenced above) and execution. With Celestia, execution happens off-chain using roll up’s that plug into the consensus + data availability layer. Applications built on Celestia use the chain only to store transactions, but not process them. Instead, the users of those applications process the transactions locally.

A great example of this model is BiTorrent. It is currently the worlds most decentralized scalable protocol because it does not do execution, it only stores and distributes data i.e. data availability.

The basic idea is that the more users join the network the more data the network can store and distribute. A key insight here is that if a blockchain can simplify or remove steps in the block verification process, it can potentially achieve similar scalability properties.

Modular Execution Layers

Layer 2’s such as( Arbitrum, Optimism, ZK Sync, and StarkWare) represent the current class of roll-up scaling solutions that lighten the load of executing transactions on-chain. The problem with these solutions is that they optimize for bandwidth only on Monolithic chains.

The beauty of a modular model is that bandwidth is unconstrained on the base layer because the layers are broken up. Celestia, specializing in data availability, opens up the amount of data the network has; and the off-chain execution layer(s) can optimize for and provide orders of magnitude more throughput(which tends to be the more important measure of network performance).

Modular execution layers are different from Layer 2’s in that they can be their own L1, L2, or L3 that provide maximum security and the highest flexible throughput for general-purpose smart contracts.

These layers, such as Fuel, run on top of Celestia, where cross-layer communication is simplified because the chains share the same data layer. This design thus reduces the negative effects of bridging. In essence, using a modular layer for execution (as opposed to roll-ups on ETH) reduces computational waste of the EVM, and can offer far more compute.

Conclusion

Over the last few years, blockchains have gone through a perpetual cycle between infrastructure and application growth. During this time, large amounts of capital and engineering resources from the public and private sectors have been allocated towards building out the infrastructure of Monolithic chains.

The reason the market is moving away from Monolithic chains is very simple. Different types of applications require different execution environments that can optimize for their needs.

The blockchain space is entering a new design paradigm, one that brings greater, by an order of magnitude, efficiency to other layers of the stack. Modularity matters because it gives users and developers adaptability, which is something they want now, and will continue to want far into the future.

This technology moves so fast that what one user may require today may soon become obsolete as the tech matures. Moreover, what one application requires today might not be sufficient in the future.

Celestia brings flexibility over execution environments to developers. But, more importantly, it allows the end-users of blockchain networks to define what they value and are willing to trade-off instead of it being defined for them.