delvingbitcoin

Second Look at Weak Blocks

Second Look at Weak Blocks

Original Postby mcelrath

Posted on: April 18, 2024 13:36 UTC

The challenges of synchronizing the mempool in a fragmented landscape, where miners accept fees out of band and thus do not relay or advertise these transactions in weak blocks, exacerbate issues for the broader network.

This practice leads to increased block validation times as other participants need to retrieve missing transactions. Weak block relay strategies are insufficient to address these concerns because they do not prevent other miners from taking advantage of the situation by sniping fees meant for those who initially received them out of band. Consequently, this approach fails to solve the underlying problem and introduces complications with the "first seen" rule of the mempool, which is crucial for constructing blocks in the presence of conflicting transactions.

The proposed solution to these challenges involves incentivizing miners to participate in decentralized mining pools rather than conducting transactions out of band. Decentralized pools offer several advantages, including lower variance, hashrate derivatives, and outsourced transaction selection, which can make it economically unviable for miners to operate outside this system. Braidpool aims to leverage these benefits to encourage miner participation within a decentralized framework.

Additionally, addressing the limitations of weak blocks requires a more robust consensus mechanism capable of time ordering and conflict resolution among transactions. Extending Nakamoto consensus to a Directed Acyclic Graph (DAG) structure, as suggested in DAGKnight and the Braid proposal, could provide a viable solution. By adding a committed header field that points to the most recently observed weak blocks, forming a DAG, one can achieve a partial order of transactions in linear time using established algorithms. This structure enables a decentralized mining pool to enforce rules ensuring that blocks contain only transactions previously broadcasted in their ancestor weak-blocks.

For further exploration of these ideas and technical details, various resources are available, including discussions on weak blocks, an overview of how a decentralized mining pool could surpass centralized counterparts, and ongoing specifications for Braidpool. These documents can be found through provided links: discussion of weak blocks, Overview of decentralized mining pool, and the WIP Spec for Braidpool.