''Dusk Does Not Treat One Good Vote as Finality."

Spent some time today afternoon going through Dusk’s consensus flow and what caught me was not how a block gets proposed. It was what happens after one committee already thinks the block is valid.

Apparently, that judgment is not enough. Long $ACE is quite risky now . $COW is on top so keep watching it also.

Succinct Attestation moves through Proposal, validation & Ratification. One committee checks the candidate block, then another independently confirms the validation result. The stake-weighted votes are aggregated with BLS signatures and ultimately become the certificate attached to the block.

Initially, i read Ratification as Dusk simply voting twice.

That misses the point.

The second round is effectively asking whether the network agrees strongly enough with the first judgment to treat the result as final. For financial settlement, that difference matters. Seeing a block is one thing. Building the next payment, transfer, or asset operation on top of it is another.

What I’m less certain about is how that extra assurance behaves when the network is having a bad day.

Suppose Validation completes normally, but Ratification participation arrives slowly. The candidate may already look perfectly acceptable, yet settlement certainty still has to wait for another committee and another round of communication.

BLS aggregation can make the resulting certificate compact. it cannot make a late participant arrive earlier.

That is probably the detail i would watch in practice.

Dusk is spending an additional round of coordination to make finality harder to mistake for simple block production. For financial infrastructure, that may be exactly the right priority.

I just want to see what that extra certainty costs when the network is least cooperative. @Dusk_Foundation $DUSK #dusk