i went back through the Babylon documentation last night because I wanted to understand the engineering decisions instead of relying on short summaries. One section that surprised me was the cost analysis. My reading is that an on-chain bandwidth case could cost around $2.66, while proof verification is estimated at roughly $0.19. What really stood out was the estimate that proof generation can take around 20 minutes on a single core and cost about $2.43. The document also suggests these costs can be reduced by amortizing proofs across multiple circuits, which I hadn't thought about before.
I then spent time on the comparison between trusted lending using a long-term custodian, trusted lending using 1-of-n multisig, and a trustless vault. The trust assumptions change depending on whether you're depositing BTC, borrowing, withdrawing collateral, or dealing with liquidation. Seeing those differences side by side made the trade-offs much easier to understand.
The later sections on Discrete Log Contracts, the BVM Bridge, and the roles of operators, signers, challengers, and challenger bonds raised even more questions for me. I understand the direction, but I'm still not completely sure how these assumptions hold up under heavy usage or how governance decisions could affect long-term decentralization.
Did anyone else read these sections differently? Which security assumption deserves the most scrutiny in your opinion?
@BabylonLabs_io $BABY #BABY