Untangling the BlockDAG
summary
Compares specific block DAG designs (GHOST, inclusive/FruitChain, parallel-chain, SPECTRE, Phantom) and presents Spacemesh as a built layered-DAG consensus: blocks tied to proofs-of-spacetime, fixed layered mesh timing, blocks point to preceding layers, and ordering is by layer then ID. Consensus uses per-block votes on validity plus a weak common coin to break balancing attacks, aiming for race-free incentives and a self-healing mode; requires stronger honest-majority assumptions than Nakamoto but removes Poisson lotteries and shows practical implementability.
AI-written summary · about
chapters (17)
- 0:00Intro
- 2:07Ledger Requirements
- 5:22Challenges for Ledger Consensus
- 7:52Consensus is Hard
- 11:23Nakamoto's Magic Trick
- 13:12Generalized Blockchain (Nakamoto Style)
- 14:51Ledger Topology
- 19:35What's Wrong With Chains?
- 23:19DAG to the Rescue?
- 24:41(Partial) Directory of BlockDAGS
- 25:45GHOST
- 27:15Inclusive Blockchains / FruitChain
- 33:52SPECTRE
- 37:44PHANTOM
- 39:55Robust to failures
- 43:04Race-Free
- 45:04Spacemesh Consensus
related talks
47mCant we just agree
22mOptimizing network topologies...
42mBeyond Blockchain: Convergent Consensus
20mShared Security in the BLOCKCHAIN
43mAll Together Now: An Introduction to Distributed Consensus
52mBinding Blockchains Together with Accountability
19mLiDO-DAG: A Framework for Verifying Safety and Liveness of DAG-Based Consensus Protocols
13mMulti-chain governance: How can DAOs live across chains
