Status Quo: L1 (EIP-4844 + ES Contract) + L2 ES Network

L2 ES network watches BLOB events on L1 ES contract and then downloads BLOBs from L1

Pros:

• Synchronous submission without data bridges

Cons:

• Gas cost is high

Solution 2: L1 (EIP-4844) + L2 (ES Contract) + L3 (ES Network)

L2 ES contract will do inclusion EIP-4844 BLOB hash check on recent (8194, ~27 hours < 18 days) L1 blocks in putBlobFromL1() method

• Optimization: using ZK to compress the proof

L3 ES Network watches BLOB events on L2 ES contract and then downloads BLOBs from L1

Pros:

• Gas saving using L2 to maintain the index of BLOB hash

Cons:

• Asynchronous submission with two transactions
• Need liveness assumption on L2 (Solution: if the L2 clock is too late compared to L1, then such L2 transactions may be invalid)
• Create challenge to ES downloader cache since L2 finalization time can be very long (e.g., including Optimism fraud proof window like 7 days)

Solution 3: L1 + L2 (EIP-4844 + ES Contract) + L3 (ES Network)

L3 ES Network watches BLOB events on L2 ES contract and then download BLOBs from L2

• Need a DA (DP) of L2
• May use Ethereum as L2 EIP-4844
  • Sequencer will submit L2 blocks with DA BLOBs to L1 (submit L2 DA BLOBs to L1 DA)
  • Maybe similar to L2 calldata with L2 EIP-4444 data expiration
• DAC (Arbitrum Nova/Metis) or DAChallange or ES DAC or bridged from Celestia or EigenDA
• Pros:
  • More gas saving that bypass Ethereum DA cost