layer-2.wiki
A structured knowledge base for Layer-2 scaling protocols
01 Rollups compress execution
Rollups execute transactions away from Layer 1, then publish enough information back to the settlement chain for anyone to reconstruct or challenge the resulting state. The scaling gain comes from batching many state transitions into one commitment.
Optimistic systems assume the batch is valid unless disputed during a challenge period. Validity systems attach cryptographic proofs that demonstrate the transition was computed correctly before the state is accepted.
Optimistic
Economic security is expressed through bonds, watchers, and fraud proofs. Latency appears at withdrawal time because incorrect assertions must remain contestable.
Validity
Succinct proofs shift trust from social vigilance toward verifiable computation. The cost profile moves into proving infrastructure and circuit design.
01 batchHash = keccak(transactions)
02 newRoot = execute(previousRoot, batch)
03 publish(batchHash, newRoot, proof)
02 Bridges are state machines
A canonical bridge records deposits, messages, exits, and finalized withdrawals. It is the accounting membrane between execution environments, not merely a token transporter.
Bridge safety depends on the finality assumptions of both layers. Fast liquidity bridges may hide latency through market makers, but canonical exits preserve the protocol's actual security boundary.
03 Data must be retrievable
Execution can move off-chain, but verification requires data. If transaction data disappears, independent reconstruction becomes impossible and the system degrades into an operator promise.
Modern designs separate execution, settlement, consensus, and data availability into explicit layers. This modularity creates a precise vocabulary for comparing systems.
Calldata
Simple, expensive, and directly available to Ethereum nodes.
Blobs
Temporary data space optimized for rollup batches and sampling.
External DA
Independent availability networks trade direct settlement-chain publication for throughput and specialized verification assumptions.
04 Specialized systems remain instructive
State channels, plasma chains, and validiums reveal the design space around custody, liveness, data, and exit games. They are not historical footnotes; they are precise boundary cases.
| System | Core assumption | Typical use |
|---|---|---|
| Channel | participants stay online | repeated payments |
| Plasma | exit data availability | asset-specific chains |
| Validium | external data committee | high-throughput apps |