Haskell Taketh Away - Limiting Side Effects for Parallel Programming

Explore the intersection of Haskell and C++ in parallel programming through this conference talk from C++Now 2017. Delve into how limiting side effects can enhance parallel programming capabilities, examining various libraries across languages. Discover the role of Haskell's purely functional approach in practical programming and its impact on software verification. Learn about deterministic parallelism, transactional memory, and accessing remote data sources. Investigate the concept of purity in programming and how entire applications can become deterministic functions of their inputs. Gain insights into the mutual influence between C++ and Haskell, their evolution, and their strengths in data baking. Understand the speaker's background in functional programming and his work on increasing safety and parallel performance in high-level, declarative programs.

Intro
Outline
Syntax quibbles
Composing first-class actions
A research project that escaped the lab
Lessons transferred / cross fertilization
Today's main theme
Promises, promises, programmer promises
Custom computation types
Custom computation: Haxl Example
Haskell built-in computation types
The separate Kingdoms of the heap
Obsessive safety, trusted code boundaries
First, lock down main
Analyze Alias-freedom
PLDI' 14Preserve Alias-freedom
Noninterference
Haskell data pros & cons