Function-Passing Style Typed, Distributed Functional Programming
summary
Prototype implements a Function-Passing model in Scala using a pickling serializer, Netty for networking, and typed, serializable closures called spores applied to stationary distributed data called silos. Three primitives — apply (lazy), send (eager, returns Future) and flatMap/shuffle — build DAG lineages like Spark for recomputation and fault tolerance. Prototype runs as an experiment (requires same class files on nodes), speeds up serialization and prevents accidental closure capture, but is not production-ready and lacks debugging stack-trace plumbing.
AI-written summary · about
chapters (5)
- 0:00Start
- 3:16been working on
- 17:54What do spores look like? Basic usage
- 19:28A Spore Guarantees...
- 27:40in SUMMARY
related talks
44mSpores: Distributable Functions in Scala
26mYo Dawg, Heard You Want To FlatMap Your Direct-Style
34mPatterns of a pragmatic functional architecture
01h 00mWhat The Functional Programming
01h 32mDistributed functional programming
30mTroll Driven Development: Implementing Functional Effects by Way of Self-Lobotomy
36mHow Functional Programming Can Make Life Easier
26mFusing Session-Typed Concurrent Programming into Functional Programming