Author: Peter Todd 2015-06-26 19:08:07
Published on: 2015-06-26T19:08:07+00:00
The discussion on Bitcoin-dev mailing list revolves around the centralizing effect of miner connectivity and how it affects block propagation. Pieter Wuille's simulations showed that miners with poor connectivity are negatively affected by other miners creating larger blocks, but Gavin Andresen argues that this effect is significant only if they do nothing to work around it. He suggests renting a server on the other side of the bottleneck and creating blocks that propagate quickly on both sides of the bottleneck. However, Peter Todd counters this by stating that "just rent a server" forces miners into deploying insecure hosted infrastructure that's vulnerable to hacking and seizure. This could lead to centralization, an effect that is already being seen as more miners sell their hashing power to centralized miners. In response to the idea that having lots of variables in the profitability equation is a decentralizing force, Peter Todd argues that as mining and hashing can be separated, this theory does not work. He asks for concrete proposals that would work against centralization of mining control. The relay network and p2pool, which are already implemented, require cooperation. However, in the long term, the p2p protocol will evolve to incorporate these optimizations, so it will require no cooperation. The co-operation comes from the fact that mempool policies have to be synchronized, not the protocol itself.
Updated on: 2023-06-10T00:29:43.786660+00:00