Ismail Khoffi
Ismail Khoffi
> Is it that relayers can cause Celestia validators to validate arbitrary block headers and signatures from other chain? Yes, though as the validator sets (or at least the +2/3...
An alternative we could consider here is to instead "white-list" the token denoms that go in and out. Basically any Tx that is not related to the celestia native token...
> It could be not part of the block in the same way the last commit doesn't need to be part of the block The last commit can't be re-computed...
Swamp PR was merged now but at least some of this was rather about e2e tests with more realistic networking instead of integration tests and mocked networks. But feel free...
A light client will need all 3 to verify a header: Header, ValidatorSet, and Commit. One could argue that everything besides the header can also be requested (on the p2p...
I've put this in the MVP project board. Although we don't need this fro implementing the MVP, this should be clarified in all detail _before_ we start the implementation (devnet).
A few questions high-level questions first and then some practical / implementation-specific ones: (1) Why is the first approach called "Subnet-based"? Is this because the mentioned topics (e.g. 0-10K) are...
Looping in @adlerjohn, @Wondertan (for some practical libp2p considerations), and @mattdf as part of the specs/applied research team too.
> Yes. It assumes only one network with one "peer list". Practical concern: do we think the whole network's peer-list (including the mapping to the "topics") fits into RAM (of...