I’ve been working on the documentation for IBC / cross-chain lately, and the more I read, the more I feel that “one-time cross-chain” is absolutely not as simple as just pressing a button… Honestly, what you end up trusting is a lot: the source chain shouldn’t roll back by itself, the light client / validator set should not act up, relayers—these messengers—shouldn’t tamper with things, and the inter-chain message channel / timeout logic shouldn’t be written sloppily; if it’s a bridge, then you also have to add the people-and-code combo of multi-signatures / oracles / escrow contracts.

Before I start taking action, I’ll take a quick look at the relevant contracts and the relayer wallet’s behavior, but it’s also kind of awkward—recently, someone has been complaining that the labels on on-chain data tools lag behind and could even mislead people… So I don’t dare to treat “label = truth”; at best, I take it as a clue. Anyway, when it comes to cross-chain, it feels like choosing a risk bundle you can live with across different trust models. What about you?