Bozo
64 posts


App Wave 4 is now live on Terminal Seven new apps are ready to be explored. Intros in the thread below:

proud to introduce @papertrade_xyz - a fair-launched, fully-onchain perpetuals exchange built on hyperliquid by @izebel_eth & @blurr -1000x leverage -0 slippage -No funding costs -Self-bootstrapping LP coming soon. learn more at: docs.papertrade.xyz







MIP-10 - Deterministic RaptorCast - has been drafted by @category_xyz researchers. This is an important upgrade to how Monad propagates blocks, and lays the foundation for faster block times and other protocol improvements. What's changing Today, RaptorCast lets a leader fan out a block as encoded chunks across the validator set, with any sufficient subset enough to reconstruct. It's already the best-in-class solution for propagating blocks to a decentralized network. MIP-10 makes it deterministic. In v0, chunks were organized into multiple Merkle trees of 32 chunks each, with each tree independently signed by the leader. v1 replaces this with a single global Merkle trie containing all of the chunks, providing a unified commitment that binds every chunk to a single canonical encoding of the proposal. In v1, the leader derives the Raptor encoding from a canonical seed = hash(round, leader, proposal time). Every chunk's position in the Merkle tree is fixed by this seed, and validators can recompute it themselves. The first valid (Merkle root, signature) pair a validator sees for a round becomes the binding EncodingCommitment. Why this matters Under the current scheme, a single Merkle root could be consistent with multiple distinct payloads. Voting on the root before decoding wasn't safe. With deterministic encoding, a validator can verify any single chunk and vote on the root immediately. That unlocks the headline benefit: consensus voting can be pipelined directly with block propagation. Validators no longer wait to fully reconstruct the block before casting their vote. One full message delay gets shaved off the critical path, paving the way for protocol upgrades that interleave consensus and meaningfully shorten block times. It also closes two attack surfaces from Byzantine validators that could have moderate effects on network health. This change can be done relatively seamlessly. There are no changes to the QC format, the voting protocol, or execution. The only additions are a canonical seed, a global Merkle root, and per-round equivocation tracking. This improvement extends the team's excellent RaptorCast work without disturbing it. The net effect is a tighter, safer dissemination layer, and the groundwork for a more pipelined consensus mechanism. Spec + reference implementation linked in the next tweet.














