Should there be a "Network Data" category?

Should there be a "Network Data" category?

Original Postby ajtowns

Posted on: January 5, 2024 07:38 UTC

In the process of optimizing network protocols, a thorough understanding of network behavior is essential.

This understanding can be enhanced by incorporating measurements or observations into the protocol design phase. Doing so acknowledges that empirical data about network performance, such as detailed benchmarks and performance analysis, are crucial for refining protocols.

For instance, when considering specific aspects like Initial Block Download (IBD) synchronization or secp256k1 signing, the question arises as to where such analysis would be most beneficial. While there's an argument to be made for including this data in the implementation section—since it directly influences the enhancement of particular implementations—it could also be advantageous to position it alongside other data analysis efforts. The latter approach would facilitate a more holistic view, allowing for cross-analysis and potentially revealing broader insights into network behavior which could inform multiple implementations or the design of new protocols.

Ultimately, the decision of categorization should support the overarching goal of protocol optimization, balancing the need for immediate implementation improvements with the broader objective of advancing the field through comprehensive data analysis and research.