Post-clustermempool package RBF: per-chunk processing

Post-clustermempool package RBF: per-chunk processing

Original Postby ajtowns

Posted on: May 23, 2024 22:40 UTC

The email discusses a nuanced aspect of Replace-By-Fee (RBF) mechanisms, specifically focusing on potential pinning issues that arise when solely relying on feerate diagram comparisons for RBF operations.

The sender identifies two primary concerns: the first being a pre-existing issue where a transaction with a lower feerate (E) would not be replaced by one with a higher feerate (V), and the second, a more concerning scenario where a high-feerate transaction (V) would be replaced by a low-feerate one (E), which contradicts the intended design of RBF rules. The latter is highlighted as a significant issue, suggesting that while some pinning scenarios are accepted due to deliberate decision-making processes, there exists a risk of unintended pinning due to complex transaction behaviors leading to the eviction of high-feerate transactions in favor of lower-feerate ones.

To address these concerns, the sender proposes several solutions aimed at enhancing the efficiency and effectiveness of RBF operations. These include the adoption of smarter RBF behavior that involves a deeper analysis to recognize superior transaction linearizations, suggesting that a better understanding and rearrangement of transactions could prevent undesirable replacements. Additionally, the concept of smarter linearization by miners is introduced, which encourages miners to invest more effort into optimizing the order of transactions within blocks, thus potentially discovering more efficient arrangements before encountering problematic transactions.

Furthermore, the idea of distributed linearization is presented, advocating for the propagation of optimized transaction sequences over peer-to-peer networks. This approach aims to leverage the collective efforts of nodes to identify and disseminate improved transaction configurations rapidly, thereby minimizing the chances of suboptimal pinning due to late transaction broadcasts. Lastly, the email mentions the implementation of small transaction relay contracts (TRUCs) as a means to limit the fee impact of low-feerate transactions by restricting their size, thereby making it easier for higher-feerate transactions to outcompete them without significant fee increases. These proposals collectively aim to mitigate the risks associated with transaction pinning and enhance the overall functionality and reliability of RBF mechanisms.