Programming Distributed Systems
summary
A concrete claim: a new language and compiler called mixy enforces that “strong outputs must stay strong” with an information‑flow type system so improper influence between weak and strong consistency is a compile‑time error. Implementation is a shim over existing databases—about 15,000 lines of C++—and the team builds a monotonic replicated system Duratio that achieves stable message delivery and reports being ~10× faster than ZooKeeper, plus a region/isolation type scheme (with optional dynamic checks) that eliminates destructive data races without runtime overhead.
AI-written summary · about
chapters (7)
- 0:00Start
- 1:45Building Programming Languages for Distributed Systems
- 11:07Composing consistency: populating rank
- 20:39Reliable Observations
- 23:02Programming monotonically
- 31:53Challenge: safely releasing locks
- 35:30Circular Doubly-Linked List
related talks
01h 01mLogic and Algebras for Distributed Computing: Applications and Open Questions
46mCoordination-Free Computations
52mI See What You Mean
29mModern Database Architecture: From SQL to NoSQL & Distributed SQL
06mLightning talk: Testing Distributed Databases
59mDistributed Abstractions
29mSecure Database Games
17mJust-Right Consistency