Eclipse — The New Era For the Modular Blockchains
1) Introduction
Eclipse allows you to create modular and customizable rollups and app chains. Initially using the Data Availability Layer provided by Celestia, Eclipse will run on the Solana VM and allow any application in the Solana ecosystem to continue as an Eclipse Rollup in the future.
Eclipse, which is part of Celestia’s Modular Fellows program, has also benefited from the Solana Foundation Grant Program. In addition to these projects, the developers of Eclipse, who have also received investment from names and companies such as Polygon Ventures, Sreeram Kannan of EigenLayer, Galileo (Jack Zampolin & David Feiock of Cosmos), and Big Brain (Solana). Eclipse are also working on partnerships with projects within the ecosystem.
Eclipse, developed by Neel Salamani and Sam Thapaliya, is ready to support more than 50 projects, including Zebeci, Notifi and Friktion. Scheduled to launch in early 2023, Eclipse will launch as an Optimistic Rollup and run all nodes itself.
You can follow Eclipse on Twitter and Discord to learn more about Eclipse and become part of the Eclipse community.
2) Architecture
2.a) Network Layers for Modular Blockchains
There are four main layers defined for the modular blockchains:
- The Data Availability layer, is responsible for ensuring that transaction data is fully and clearly distributed.
- The Consensus layer, is responsible for defining the order of transactions.
- The Settlement layer, is responsible for implementing the globally accepted and agreed-upon state.
- The Execution layer, is responsible for implementing changes to the state of the network.
The stronger and more scalable the Data Availability and Consensus Layers are, the more secure the network is. Both layers must be strong.
The Settlement and Execution layers create the necessary environment for a blockchain. The execution layers that share the same settlement layer form a cluster that does not require trust.
Both layers must be strong. The settlement and execution layers create the necessary environment for a blockchain. The execution layers that share the same settlement layer form a cluster that does not require trust.
What does it mean for a cluster not to require trust?
Blockchains that work monolithically, i.e. those that operate all four layers in a single network, carry the burden of all four layers on their own. While having all four layers on a single network provides high security and ease of use, it also opens the door to high usage fees and limited scalability.
2.b) Rollup
We have all heard that Ethereum’s roadmap has changed to be based on Rollups. What does this mean?
Ethereum provides the Data Availability, Consensus, and Settlement layers for Rollup applications, known as Layer2, while execution is completely carried out on the Rollup.
Rollup chains send the transfers they execute within themselves to Ethereum in a way specific to the type of Rollup, reducing the execution load on the Ethereum main network.
Rollups built on Ethereum’s Settlement layer form a “cluster that does not require trust” because the entire Rollup status is stored on the Ethereum main network. If data transfer is desired between two Rollups, the approved status in the settlement layer can be checked.
If we are talking about an Optimistic Rollup, this type of Rollup sends all the transactions it executes within itself to the Ethereum main network and receives a type of evidence called a Fraud Proof from the Ethereum main network back to the Rollup if necessary. After the transactions are broadcast on the Ethereum network, a Fraud Proof must be generated and sent to the Rollup within a certain period of time (challenge period).
The trustworthiness of a single node on an Optimistic Rollup is enough to verify the transactions and Fraud Proof. This structure brings Optimistic Rollups to the “honest minority security model”.
There are two types of Fraud Proofs for Optimistic Rollups.
When a Non-Interactive Fraud Proof is sent to an Optimistic Rollup, the proof requires all the transactions in the block to be processed again. This process is very costly.
When an Interactive Fraud Proof is sent to an Optimistic Rollup, both the Rollup nodes and the main network nodes must be active. Then, a challenge called the “Bisection Game” is completed and it is determined which transactions the main network nodes consider invalid. The “Bisection Game” requires a third party. After the invalid transaction is determined, it is re-executed on the main chain with the “Virtual Machine Interpreter”.
If we are talking about a ZK Rollup, after the transactions on the Rollup are executed, the compressed form of the block obtained is sent to the Ethereum main network with a proof called a Validity Proof, which is also generated within the Rollup. It is not necessary for any node on the ZK Rollup to be trustworthy. A machine or circuit must be designed to process the compressed block and Validity Proof structure (probabilistically checkable proof (PCP)) generated by the Prover.
Rollup chains provide “Shared Security” because they use common Data Availability and Consensus layers. This means that rollup chains connected to Ethereum share the security of the Ethereum main network, providing high-level security.
Despite rollup chains, the Ethereum main network still has scalability issues. These discussions start due to those who claim that the Data Availability Layer of the Ethereum main network is not scalable.
2.c) Celestia
The Celestia project, which saw the deficiency at this point, provides a common Data Availability and Consensus Layers service for blockchain networks. The Celestia team, which has very well observed the shortcomings of Ethereum, which started the modular era with rollup chains, is removing the barriers to scalability as much as possible by taking the modular era to the next level.
So how does Celestia do this?
I mentioned at the beginning that the Data Availability and Consensus Layers provide security for blockchain networks. Blockchain networks built on Celestia will build new chains using Celestia’s Data Availability and Consensus layers and, together with other projects on Celestia, will provide shared and strengthened security.
Celestia, which will provide Data Availability Sampling with Reed Solomon Erasure Coding, can ensure the integrity and accuracy of any data shared with the blockchain network with evidence called Fraud Proof and Data Availability Proof.
Two new types of Rollup chains, called “Settlement Rollup” and “Sovereign Rollup” have emerged.
I previously mentioned four basic layers for blockchain networks;
→ Execution
→ Settlement
→ Data Availability
→ Consensus.
Sovereign Rollup leaves Data Availability and Consensus Layers on a chain while providing Execution and Settlement on its own.
Settlement Rollup, on the other hand, leaves Data Availability and Consensus layers on a chain while not performing any execution on its own. A chain serving as a Settlement Rollup only provides the Settlement layer for new Rollups and leaves the Execution task to its new chains.
2.d) The New Era — Eclipse
Eclipse will start its life as a Sovereign Rollup on Cosmos. The project, which will use Celestia for the Data Availability and Consensus layers, will launch the Settlement layer, which will provide honest minority level security, in early 2023. The Solana ecosystem, which can be connected to other Cosmos chains using the Inter-Blockchain Communication Protocol developed by Cosmos, will open the Solana ecosystem to the entire blockchain world.
Eclipse’s Execution Layer will be a fork of the Solana execution layer. This means that we will see Solana’s Virtual Machine, Sealevel (parallel execution tool), and Pipelining (Transaction Processing Unit / TPU) innovations on Eclipse as well.
The fork of the Solana Execution layer allows the Solana ecosystem to;
→ Enable Solana projects to open up to the multichain ecosystem and use IBC technology,
-> Give independence to Solana projects and become app-specific chains,
-> Access high liquidity, and
-> Enable the possibility of moving to their own chains in a fully customizable way.
Eclipse can also bring real-world applications such as decentralized central limit order books (CLOBs), CLOB-compatible decentralized applications, messaging, and games to the Cosmos ecosystem, which currently does not have them.
Initially launching as an Optimistic Rollup, Eclipse can continue its life as a Zero Knowledge Rollup in the future as well. The Eclipse development team is also working on the Zero Knowledge Rollup Architecture in parallel with the Optimistic Rollup architecture.
Eclipse will use the eBPF (extended Berkeley Packet Filter) virtual machine used by Solana. Eclipse, which will start its life with the Optimistic Rollup architecture, will use the Interactive Fraud Proof structure. At this point, two paths are open for Eclipse;
→ Developing a Virtual Machine Interpreter for eBPF Virtual Machine.
→ Create a zero-knowledge proof for the single transaction to be re-executed.
The Eclipse team, which is also working on a Zero Knowledge Rollup in parallel with Optimistic Rollup, aims to create a project that will support all existing Solana dApps. Therefore, it is necessary for the Eclipse team to arrange the entire eBPF Virtual Machine as a zero-knowledge circuit in order for Eclipse to work as a Zero Knowledge Rollup.
A Zero Knowledge Virtual Machine must be able to run all programs. At this stage, two options are available to the Eclipse team;
→ Develop an intermediary Virtual Machine, such as Cairo VM or Miden VM, which can read Zero Knowledge inputs,
→ Redesign an existing Virtual Machine, such as EVM or eBPF, as a Zero Knowledge circuit.
Some zk-EVM designs already exist. For the Solana execution layer, which uses an LLVM compiler structure, it is an option to develop an IR (intermediate representation) program that converts the written code into an ELF (Executable and Linkable Format) format that can be read by the ZK circuit. If the structure seems complex, let’s summarize;
— — — — — — LLVM -> IR -> ELF -> ZK Circuit — — — — — —
The Eclipse team seems to be inspired by the Scroll zk-EVM design at this point. An EFL file produced using LLVM and given as input to the ZK circuit will be given as input to the ZK Circuit and the output will be sent to the Settlement Layer.
It is becoming more clear that the Eclipse team has a very difficult development process ahead of them :)
— — — — — Author’s Comment — — — — —
When considering the Optimistic Rollup and Zero Knowledge Rollup architectures, Optimistic Rollup is a structure that requires much less workload. Technically everything is ready. However, it is harder to move Optimistic Rollup structures to another Settlement Layer.
Both Rollup architectures can be occupied and transactions can be prevented by producing fake proofs.
Zero Knowledge Rollup is slow due to the architecture because generating proof requires high processing power. In addition, there are many unanswered questions for the Zero Knowledge Rollup architecture. Technically, abandoning the Optimistic Rollup architecture without creating a flawless Zero Knowledge Rollup architecture will not yield good results for Eclipse.
Eclipse, which plans to become a Settlement Rollup in the future, will provide a Settlement layer for projects that want to run their applications on it and these projects will create their own App Chain. These app chains will be responsible only for execution and will be able to customize their own chains as they wish. These chains, called Layer3, will be able to choose their desired chain as the Data Availability layer. With this design, Eclipse L2, which will increase its scalability, will also have some small decentralized applications.
Author’s Comment
Allowing each app chain to choose its own Data Availability and Consensus layers provides freedom for app chains. However, this can cause problems for app chains that choose Data Availability and Consensus layers that are not sufficiently secure and scalable.
Data sharing between Rollups that share the same settlement layer can be done in a trust minimized manner. This allows app chains built on Eclipse to communicate with each other in a trust minimized manner. However, app chains that do not share the same Data Availability and Consensus layer will have a security gap between them.
Tokenomics is also an important issue for Rollup chains. Many Rollup chains have recently released their own tokens. Therefore, tokenomics is a very new topic for Rollup chains.
There is a list of suggested uses for Eclipse Token by the development team. Some items on the list include:
- Burning or distributing Sequencer and Prover fees to Eclipse Token stakeholders
- Giving MEV revenues to stakers
- Allocating funds to dApps built on Eclipse
- Supporting corporate application chains
- In-chain voting mechanism
- Earning yield farming revenue from Eclipse Rollups
Incentives such as airdrops, event awards, and contributor awards may also be included in this list.
3) Conclusion
The Eclipse project, which will open the Solana ecosystem to the world of blockchain and provide high scalability with its modular structure, will facilitate the expansion of the Cosmos ecosystem and the recognition of the Solana ecosystem.
By bringing the modular blockchain structure to a higher level with the Optimistic Rollup and Zero Knowledge Rollup architectures it is working on, Eclipse aims to provide a top-level experience for both users and developers by also taking advantage of some Solana developments (Solana VM, Pipelining, Sealevel). The eagerly awaited Eclipse project will be launched in the early stages of 2023 and will take scalability to a completely different level.”
4) Resources
Thank you for reading.