Author: Paul Sztorc 2017-12-05 19:55:33
Published on: 2017-12-05T19:55:33+00:00
In this email thread, Paul Sztorc, a Bitcoin developer, responds to criticisms of the security model for drivechains. The criticism is that non-verifying miners will be complicit in theft due to upvoting by default and the potential for miner centralization. However, Sztorc notes that this is not centralized since each miner can independently make decisions. In response to concerns about multiple withdrawal attempts weakening the peg, Sztorc explains that there are mechanisms in place to address these problems. Multiple withdrawals are allowed simultaneously, but only one can be upvoted at a time per sidechain per main block, which allows for an "auditor class" to ignore all withdrawals until a significant amount of time has been invested in one candidate. A "farmer class" can then take it upon themselves to make sure the good withdrawals get in and get upvotes. Sztorc agrees with the idea that miners should always accept BMM bribes, which is basically the same as saying that miners should always mine on top of the heaviest chain. However, he disagrees with the conclusion that the optimal policy is always downvoting. Even if this analysis turns out to be correct, it is not a total disaster because miners are the ones who perform atomic swaps and walk coins side-to-main. As long as there are a few large mining groups, competition will drive atomic swap fees down to negligible levels. The email thread also includes a discussion on the security model for sidechains versus drivechains. While some argue that there are substantial opportunity costs and collective action problems when it comes to rewriting the mainchain, there doesn't seem to be anything similar for drivechains. However, Sztorc notes that the security model for sidechains is the same as that for every blockchain and that criticisms of drivechains' security are unfounded.
Updated on: 2023-06-12T22:31:55.227795+00:00