Java Keynote at Oracle Code One 2018
summary
GitHub describes choosing Java for data pipelines and builds Hydro uses Kafka/HDFS/Presto/Flink to unify queues and analytics, citing Ruby’s GIL, memory cost, and lack of strong typing. demos show Git migration experiments, GitHub apps that run JCheck bots on pull requests, and Project Skara work. Java’s release model is now six‑month feature releases with LTS every ~3 years (JDK 9,10,11), and language/runtime work includes Loom fibers, Panama FFI, Valhalla value types plus matrix multiply: 3.7s → 0.3s, GC 4GB→4MB.
AI-written summary · about
chapters (32)
- 0:00Start
- 4:54Matthew Mccullough
- 14:07Emerging Messaging Systems
- 14:36Web Hooks
- 15:06Analytics Pipeline
- 15:32Project Hydro
- 16:27Parallelism
- 16:50Performance
- 17:30The Ecosystem
- 19:18Third Act
- 20:01Jdk
- 25:45The Evolution of Java
- 28:42Long Term Support Releases
- 34:11Modularizing the Platform
- 39:22Jaidev
- 42:35Api's Deprecated in Java 11
- 45:01Support
- 55:59Amber
- 56:48Local Variable Type Inference
- 58:56Raw String Literals
- 1:01:49Switch Expressions
- 1:06:09Pattern Matching
- 1:06:49Loom
- 1:08:34Context Class Loader
- 1:10:58Completable Futures
- 1:17:31Panama
- 1:18:42Demo
- 1:22:50Clock Get Time
- 1:25:37Valhalla
- 1:26:44Value Types
- 1:28:55Matrix Multiplication
- 1:31:10Instructions per Cycle
