Yea but any double signing of like two pre-confirms would just automatically go full nuclear and cause the two rollups to diverge and hard fork then
Respective assets would become worthless on the other etc
@joncharb.bsky.social
Co-founder dba.xyz | Lower-case r research
Yea but any double signing of like two pre-confirms would just automatically go full nuclear and cause the two rollups to diverge and hard fork then
Respective assets would become worthless on the other etc
Unless you're just saying the sequencer must've double signed at some point even if just at the rollup/pre-confirm level, so showing that would automatically set off both bridge alarms
02.07.2023 04:00 β π 1 π 0 π¬ 1 π 0Unless you seem to be getting at each bridge just optimistically assumes the other bridge is in line, then you'd prove a fault/divergence? (so would need a challenge window even if bridges are proving the rollup in ZK)
I don't see how to prove this fault/divergence appropriately
I don't see how it's possible to synchronize the chains/bridges here on timing, and prove that the other bridge is in line for the same time?
02.07.2023 03:51 β π 0 π 0 π¬ 2 π 0Each bridge would only be willing to update their view of the canonical tip of the rollup chain when the other bridge contract also agrees on that same new state of the chain
If they diverge, each contract has a kill switch where they fall back to only looking locally
So every rollup full node would need to run:
1) a full node of the rollup itself
2) embedded Ethereum full node
3) embedded Solana full node
You'd be posting full DA to both chains (and charging for both), and you'd have bridges on both which the rollup logic is tied to (e.g., process deposits)
Iβll go second π
02.07.2023 02:58 β π 3 π 0 π¬ 1 π 0Yea including local txs seems easy
But how would eg the Solana bridge contract know that the Ethereum bridge contract is in agreement with its view of the rollup?
How would it be possible to mirror? Each bridge contract would need to be able to check without new trust assumptions that the bridge contract on the other chain matches itβs view of the rollup at the same time
02.07.2023 02:35 β π 0 π 0 π¬ 1 π 0And you cant have the rollup be pulling from both DA layers, because then a fault on either side causes a fault on the other
02.07.2023 02:21 β π 0 π 0 π¬ 1 π 0Yea but fundamentally how would you construct the chain? Letβs say itβs a rollup that wanted to do this for Ethereum and Solana
Each bridge needs to have a notion of what the STF for the rollup is, if each one goes off itβs own DA layer, they wouldnβt be the same rollup
Could def have a rollup with a validating bridge to βsettleβ to multiple chains (rollups should start doing this imo), but will need to pick one layer for DA that the chain follows
So using the other bridges not from the DA layer will just have an added trust assumption
gm
02.07.2023 00:17 β π 7 π 0 π¬ 3 π 0