Correctness Conditions for Cross-Chain Deals
summary
Paper claims existing atomic-swap models don’t cover brokered cross-chain deals where a broker uses a buyer’s funds to pay a seller; assets (tickets and coins) live on separate blockchains and the protocol must guarantee that any conforming party is not worse off. Practical mechanisms borrow HTLC ideas and CBC variants, but atomic swap is insufficient and an escrow required to prevent mid-deal theft. Main theorem: correctness (no honest party underwater) holds iff the directed exchange graph is strongly connected. Paper lives on arXiv and will appear in VLDB.
AI-written summary · about
chapters (19)
- 0:00Intro
- 1:18What is an adversary
- 3:21The world has changed
- 5:15An adversarial commerce
- 6:57A naive abstract deal
- 9:08A swap
- 10:37What can go wrong
- 12:32Identity
- 18:44Classical notions of correctness
- 19:58Acid properties
- 21:05Failure model
- 21:28Compliance model
- 26:06Allornothing properties
- 32:01Game theory
- 33:16Different payoffs
- 36:30Deal is better
- 39:02Underwater outcomes
- 43:41Strong Nash Equilibrium
- 51:43Conclusion
related talks
15mCrossing Chains, Losing Gains
27mSmart Transactions
27mWar and Peace: The Multi-Chain Reality
31mSending Cross Chain Messages
16mExploring the Cross-Chain Interoperability Protocol (CCIP)
17mThe Future of Interchain Composability
42mBeyond Blockchain: Convergent Consensus
56mA sledgehammer to crack a nut: why blockchain is not (always) a good idea