#dusk $DUSK @Dusk Many people who look at Dusk's consensus layer ask one question first - how fast is it? I kept thinking that speed is the wrong place to start. The more interesting question is what the design is actually protecting against.
Succinct Attestation runs every block through three steps - Proposal, Validation, Ratification. A provisioner creates a candidate block. A randomly selected committee checks its validity. A second, separately selected committee confirms the validation outcome and finalizes it. Two independent committees, same block, same round.
What slowed me down reading this was the word "independently." Most PoS systems use one committee to propose and approve. Dusk uses two separate committees for validation and ratification, meaning a block needs agreement from two groups who were selected randomly and separately before it is ever considered final. A single compromised committee cannot push through an invalid block alone.
The randomness matters more than it first appears. Provisioners are selected through sortition - weighted by stake, but non-deterministic. Nobody knows in advance which specific nodes will form the next validation or ratification committee. That uncertainty is the security mechanism. An attacker cannot bribe or target a committee that does not yet exist.
The minimum stake to participate is 1,000 DUSK. That number is not arbitrary - it sets the economic floor for what it costs to become a provisioner and potentially influence committee selection. The higher the total staked, the more expensive it becomes to accumulate enough weight to matter.
What I find genuinely unresolved is the relationship between stake concentration and sortition outcomes over time. If a small number of large provisioners dominate the stake weight, the "random" selection starts producing predictable patterns. The cryptography holds, but the economic assumptions underneath it require watching.
Succinct Attestation runs every block through three steps - Proposal, Validation, Ratification. A provisioner creates a candidate block. A randomly selected committee checks its validity. A second, separately selected committee confirms the validation outcome and finalizes it. Two independent committees, same block, same round.
What slowed me down reading this was the word "independently." Most PoS systems use one committee to propose and approve. Dusk uses two separate committees for validation and ratification, meaning a block needs agreement from two groups who were selected randomly and separately before it is ever considered final. A single compromised committee cannot push through an invalid block alone.
The randomness matters more than it first appears. Provisioners are selected through sortition - weighted by stake, but non-deterministic. Nobody knows in advance which specific nodes will form the next validation or ratification committee. That uncertainty is the security mechanism. An attacker cannot bribe or target a committee that does not yet exist.
The minimum stake to participate is 1,000 DUSK. That number is not arbitrary - it sets the economic floor for what it costs to become a provisioner and potentially influence committee selection. The higher the total staked, the more expensive it becomes to accumulate enough weight to matter.
What I find genuinely unresolved is the relationship between stake concentration and sortition outcomes over time. If a small number of large provisioners dominate the stake weight, the "random" selection starts producing predictable patterns. The cryptography holds, but the economic assumptions underneath it require watching.