[BIP Proposal] Stale Tip Relay

Introduced as an optional peer-to-peer (P2P) messaging system, staletip allows nodes to share information about recently observed stale tips with other peers that have opted into this feature. This initiative builds upon AJ Towns' foundational work and has advanced through collaborative efforts between pseudoramdom and w0xlt, who have also created a proof-of-concept available for review (Draft BIP, Proof-of-concept).

Stale blocks, which occur when two valid blocks are found almost simultaneously at the same height, present challenges in accurately monitoring and understanding their causes due to current relay technologies like compact block relay and FIBRE that prioritize rapid propagation of the longest chain. This often results in stale branches being unnoticed unless closely monitored, and even then, monitors may only see parts of the stale branches. The elevated stale rates could signal various underlying problems such as network partitions or adversarial mining strategies, making it crucial to have tools like staletip for better diagnostics and quicker reorganization handling.

The technical specifications of staletip include the use of BIP 434 feature messages for nodes to indicate support for this function. The messages related to staletip contain details such as the fork point, a sequence of compressed headers, and a status flag indicating whether the sender can provide the stale tip block. The policy proposed limits the communication to tips within 1000 blocks of the active chain tip to avoid outdated or irrelevant data and restricts the header sequences to 20 compressed headers to manage the size and scope of information shared. Moreover, each node maintains a cache of up to 10 recent tips, ensuring efficient data transmission without overloading the network.

By introducing staletip, the developers aim to provide a system where nodes are better informed about network events, potentially leading to improved response times in active branch switches and deeper insights into the dynamics of block propagation and stale rate occurrences. Feedback on this proposal is encouraged to refine and optimize its implementation further.