Latency and Privacy in Lightning

The discussion begins with a challenge to the assertion that routing nodes in payment networks are primarily motivated by a desire to minimize latency. The counter-argument presented highlights that the main performance bottleneck for these nodes is actually disk IO, specifically database operations required when commitment_signed messages are sent. This operational necessity leads nodes to adopt a strategy of batching HTLCs (Hashed Time-Locked Contracts) to reduce the frequency of disk IO, which is particularly crucial for nodes that handle a high volume of payments. Batching not only mitigates the disk IO issue but also results in more stable latency times for the network. Although this approach increases median latency, it benefits the system by reducing the variability of latency times, optimizing the overall performance and user experience on the network.

Furthermore, the debate touches upon the expected behavior of nodes if the Lightning Network succeeds as a payment platform. The emphasis is on finding a balance between user experience and node performance, suggesting that an optimal batching interval can be achieved without significantly compromising payment processing times. The conversation also delves into privacy measures, where it's suggested that incorporating receiver-side and sender-side random delays, alongside randomized batching intervals at intermediate nodes, could enhance privacy without detrimentally affecting network performance. Additionally, the use of random message padding or cover traffic is proposed as a method to improve privacy protections further.

The discussion concludes with an open question regarding the necessity and effectiveness of intermediate forwarding delays for privacy, even in the presence of other privacy-enhancing techniques such as random message padding and cover traffic. This inquiry reflects a deeper curiosity about the balance between privacy and efficiency within the network's operation, inviting input from other experts in the field to enhance the understanding and implementation of these mechanisms within the Lightning Network.