Optimal library architecture for large-scale formalization

Ascertain whether a single monolithic foundational library (or a bifurcation into separate mathematics and physics libraries) is the best and scalable architecture for formalization across the mathematical sciences, or whether multiple specialized libraries with shared standards provide a more effective and sustainable approach.

Background

The paper discusses the bottlenecks in integrating formalizations into a central library such as Mathlib and the risk of fragmentation as independent projects define overlapping concepts differently.

The authors raise the question of whether a single monolithic library is the right long-term solution for broad formalization across mathematics and physics, suggesting alternative models (e.g., specialized libraries coordinated by shared standards) might be more scalable.

Determining an optimal architecture has implications for interoperability, contributor onboarding, review bandwidth, and long-term maintainability of formalized knowledge.