Interview
summary
Develops promise as an atom of policy to model distributed systems and CFEngine behavior, claiming promises are voluntary, non‑guaranteed commitments that enable resilient cooperation unlike a brittle obligation model. Applies dimensional analysis to identify scale‑pinning software constructs (for example tight loops, single‑process state, or layered coupling) and argues web‑scale wins by parallelized microservices. Proposes a future software wind tunnel for scaling experiments, estimates it a decade away, and says current tools are mainly profilers that reveal but cannot yet fix scale limits.
AI-written summary · about
