Just finished going through a round of trust models for IBC versus ordinary bridges, and there are some things I want to write, but I feel like no one will read it. Forget it.

To put it simply, the hardest part of cross-chain isn’t the technology—it’s who you actually trust. For an ordinary bridge, you basically hand your coins over to a multisig group, then they “map” you a balance based on their mood. In essence, it’s credit custody.

IBC is a bit better: the set of validators is shared, consensus security is tied to the base chain, and an attacker can’t simply “change state” on either side without involving the validators—because they’re directly able to alter state only as validators tied to the relevant chain, as long as both ends use the same consensus and don’t go through relay forks.

So what about relayers? Relayers themselves don’t create security—they just do work, relay messages, and don’t control your assets. So if you run through multiple independent relays, the attack cost is much higher than with ordinary bridges.

Recently I’ve been seeing a lot of testnets people farm points on, and at the same time gamble on new token issuance. Some projects’ bridges are actually built pretty well, but the moment tokens get issued, I get scared—will it turn into something like “change parameters to pump the price”? I’m not making wild guesses, but you should take a close look at whether their IBC monitoring dashboards expose on-chain data publicly.

Anyway, I’ll set up two layers of routing first—slower, but steady.