Core.Async in Use
summary
Core.async often gets used at the wrong level: channel I/O is sprinkled through business logic, causing non‑determinism and hard-to-find bugs like a 500-deep concat stack that killed performance. Prefer transducers and pipeline helpers to express batching and rebuffering declaratively (pipeline with transducers or attach transducers to channels), expose async results via callbacks for libraries, and contain async with synchronous interfaces such as Om Next queries or reified call stack interceptors so daily UI and domain code stays pure and testable.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 2:24The Problem
- 11:29Return types
- 12:44Patterns
- 16:22Interesting Things
- 19:00Autocomplete
- 21:49Reification
- 26:36Functional Reactive Programming
- 30:12Actors
related talks
01h 04mcore.async: Concurrency without Callbacks
40mCore.Async
44mClojure core.async Evolution
35mApplying the paradigms of core.async in ClojureScript
43mClojure in live sports television
30mRealtime Collaboration with Clojure
21mGaticos & Monads
28mFrom Clojure to Elixir: Leveraging Functional Programming for Scalable...