I went back through the Dusk whitepaper last night, especially the sections on incentives, transactions, and Moonlight, and I found the design more nuanced than I first expected.

The consensus side uses 64 committee credits, with voting power weighted by credits. Quorum needs 2/3 for Valid, while Invalid, NoCandidate or NoQuorum can pass with 1/2 + 1. Rolling finality also caught my attention: if a block has two previous non-attested iterations, it needs 2×2 = 4 consecutive attested or confirmed blocks to become confirmed.

The incentive model is interesting too. Block rewards are split 80% to the generator, 10% to the voting committee, and 10% to Dusk. The generator’s 80% itself includes a fixed 70% plus a variable 10% tied to included votes. I can see why this exists: higher-iteration generators could otherwise benefit from earlier iterations failing.

On transactions, Moonlight is account-based and transparent, with public balances and a nonce for replay protection. Phoenix takes the UTXO route and uses ZK proofs and nullifiers for privacy.

What I’m still wondering is whether the 80/10/10 structure creates enough incentive for broad participation, and how decentralization behaves when stake concentration determines committee credits.

How does the community view these trade-offs?


#dusk $DUSK @Dusk