Conversation

(just have the program filter out invalid chains and then choose the longest remaining) So the "consensus" mechanism here is "if any single person is honest then the right thing happens" which is pretty sweet. E.g. you yourself can guarantee it by submitting the real history.
2
2
So the likely answer is: a) things before are fine b) tx's during the fork are messy af c) once the dust settles you create a new bridge program if needed and people transition over (anyone can create one but people will migrate to the right one unless there's a fork).
1
1
Yea, I'm generally concerned about validator sets from non-ETH chains being incentivized to steal assets locked in ETH bridges. If the txns sent during the fork are more valuable than the SOL rewards for the fork period (including slashing penalties) then it might be worth it.
1
3
You could also imagine people on the ETH side rejecting/feezing/etc. assets if they were stolen this way. Mostly I think it's not very likely, but it's def important to make sure it's not too out of whack. Also prob SOL is worth way more in the world where it's this important.
1