4 ms·

cross chain bridges, from my understanding, are layer 2 protocols that live on 2 chains at once using validators. Basically, you have nodes staking coins on bot

cross chain bridges, from my understanding, are layer 2 protocols that live on 2 chains at once using validators. Basically, you have nodes staking coins on both chains and they attest that the coins coming out of one side of the bridge are matched by coins going in the other side of the bridge So for example you send some $FOO tokens to a contract on the ethereum side then get out $WFOO (wrapped foo) tokens from a contract on another chain. I've been interested in these protocols for a while as a possibility of making crypto actually scalable and not a pyramid scheme (basically, imagine a network of interoperable chains powered by stable tokens like USDC, and using inter-chain protocols to transfer USDC between them as needed for load balancing). That said, reality hasn't quite met my expectations, not yet anyways.

p4bl0 4y ago
What makes a blockchain necessary is having both decentralization and generalized distrust as hard assumptions. How do you manage not having to trust the bridges in such scenarios?
- helen___keller 4y ago
  the bridges operate on the same general principle as proof of stake (mutual distrust of nodes through a protocol that involves staking coins) The difference of course, is that if there's a flaw in proof of stake generally it means one person might control transaction flow and collect some transaction fees; if there's a flaw in a bridge then someone prints money and destroys bridge ecosystem
  - smoldesu 4y ago
    > the bridges operate on the same general principle as proof of stake Quick amendment here - most of the bridges operate on these principles. There are hundreds of cryptocurrencies, trading pairs and L2 chains that have zero accountability whatsoever. It's not written anywhere that these bridges have to operate without trust, and many of them straight-up don't.
  - p4bl0 4y ago
    Proof of stake comes with a number of issues but let's put that apart. PoW and PoS are means to achieve distributed consensus. It works for cryptocurrencies because writings to the blockchain are performative (what's written is true because the act of writing it makes it true). This can only work inside a given blockchain and only for its own cryptocurrency. Which means it cannot be used to ensure anything that happens outside that blockchain. In particular all blockchain usage for certifications, traceability, of NFT-like things are essentially nonsensical. Similarly, I don't see how it could work for writings that happens on other blockchains. Whatever is written on blockchain A cannot be considered true in blockchain B without breaking the security model of blockchain B by making blockchain A a trusted third-party. Where does a distributed consensus mechanism comes into play here? Maybe it would be clearer with an example? Let's say someone wants to move X tokens from blockchain A to blockchain B, and that at the time of the move, X A-tokens are equivalent to Y B-tokens (trusting the exchange rate in itself is all another problem, but let's put it aside). How would that go?
    - helen___keller 4y ago
      > Whatever is written on blockchain A cannot be considered true in blockchain B without breaking the security model of blockchain B by making blockchain A a trusted third-party. To be more specific, when I said "like proof of stake", what I meant is that there's a trusted pool of validators but the individual validators are not assumed to be trustworthy (only the pool in aggregate), and the validators use a adversarial staking model to punish dishonesty. In the case of bridges, that pool is external to the chain. So yes, there is external trust happening. I'll try to address your example, but I've researched this a while ago so I'm sure I could be off the mark. For your example, lets say we're moving A-token in A-chain to wrapped A-tokens in B-chain, so we can easily assume it's a 1:1 exchange rate by construction. User moves 100 A-tokens in A-chain to a bridge smart contract and as data passes the address of their B-chain wallet. A subset of validators in the trusted pool (chosen via some distributed protocol) together sign a multi-sig transaction that authorizes the bridge contract in B-chain to mint 100 wrapped A-tokens and send 99.9 of them to the destination address that the bridging user, and collect 0.1 of them as fee. (If transaction were dishonest, a larger subset of validators would together sign a transaction purging those validators' stake in the pool; so the subset of validators chosen are honest out of desire to keep their staking capital and earn the 0.1 fee) That's at a high level how it works. If you want the full details I recommend finding a technical writeup for a popular trustless bridge.

p4bl0 4y ago

What makes a blockchain necessary is having both decentralization and generalized distrust as hard assumptions. How do you manage not having to trust the bridges in such scenarios?

helen___keller 4y ago
the bridges operate on the same general principle as proof of stake (mutual distrust of nodes through a protocol that involves staking coins) The difference of course, is that if there's a flaw in proof of stake generally it means one person might control transaction flow and collect some transaction fees; if there's a flaw in a bridge then someone prints money and destroys bridge ecosystem
- smoldesu 4y ago
  > the bridges operate on the same general principle as proof of stake Quick amendment here - most of the bridges operate on these principles. There are hundreds of cryptocurrencies, trading pairs and L2 chains that have zero accountability whatsoever. It's not written anywhere that these bridges have to operate without trust, and many of them straight-up don't.
- p4bl0 4y ago
  Proof of stake comes with a number of issues but let's put that apart. PoW and PoS are means to achieve distributed consensus. It works for cryptocurrencies because writings to the blockchain are performative (what's written is true because the act of writing it makes it true). This can only work inside a given blockchain and only for its own cryptocurrency. Which means it cannot be used to ensure anything that happens outside that blockchain. In particular all blockchain usage for certifications, traceability, of NFT-like things are essentially nonsensical. Similarly, I don't see how it could work for writings that happens on other blockchains. Whatever is written on blockchain A cannot be considered true in blockchain B without breaking the security model of blockchain B by making blockchain A a trusted third-party. Where does a distributed consensus mechanism comes into play here? Maybe it would be clearer with an example? Let's say someone wants to move X tokens from blockchain A to blockchain B, and that at the time of the move, X A-tokens are equivalent to Y B-tokens (trusting the exchange rate in itself is all another problem, but let's put it aside). How would that go?
  - helen___keller 4y ago
    > Whatever is written on blockchain A cannot be considered true in blockchain B without breaking the security model of blockchain B by making blockchain A a trusted third-party. To be more specific, when I said "like proof of stake", what I meant is that there's a trusted pool of validators but the individual validators are not assumed to be trustworthy (only the pool in aggregate), and the validators use a adversarial staking model to punish dishonesty. In the case of bridges, that pool is external to the chain. So yes, there is external trust happening. I'll try to address your example, but I've researched this a while ago so I'm sure I could be off the mark. For your example, lets say we're moving A-token in A-chain to wrapped A-tokens in B-chain, so we can easily assume it's a 1:1 exchange rate by construction. User moves 100 A-tokens in A-chain to a bridge smart contract and as data passes the address of their B-chain wallet. A subset of validators in the trusted pool (chosen via some distributed protocol) together sign a multi-sig transaction that authorizes the bridge contract in B-chain to mint 100 wrapped A-tokens and send 99.9 of them to the destination address that the bridging user, and collect 0.1 of them as fee. (If transaction were dishonest, a larger subset of validators would together sign a transaction purging those validators' stake in the pool; so the subset of validators chosen are honest out of desire to keep their staking capital and earn the 0.1 fee) That's at a high level how it works. If you want the full details I recommend finding a technical writeup for a popular trustless bridge.