Changing the Unchangeable: The Hows and Whys of Immutable Data Structures

on YouTube ↗35m

" summary

Demo shows a Bloxorz solver and a Connect Four UI built on a Hamster-backed matrix gem and an Elm frontend, with the solver implemented as a breadth-first flat_map tree search. Persistent lists are singly-linked; vectors use a persistent bit-partitioned trie that shares nodes so updates copy only a few nodes. These structures are lazy and thread-safe, give cheap diffs and fast immutable updates, and enable time-travel debugging workflows (Redux + Immutable.js, Datomic-style history, hot-reload rewind shown working).

AI-written summary · about

" chapters (9)

1. 0:00Intro
2. 5:22Persistent Data Structures
3. 11:14Boxers
4. 14:50Tree Search
5. 18:01Lazy Algorithms
6. 22:37Demonstration
7. 26:18Demo
8. 28:21Discussion
9. 30:54Elm

" related talks

• 26mImmutable RubyMichael Fairley · MountainWest RubyConf 2013
• 39mImmutable RubyMichael Fairley · RubyConf AU 2013
• 59mImmutable Data Structures in .NET: Why They Matter and How They WorkDmitriy Ivanov · .NET Fest 2018
• 48mA Helicopter Tour of Purely Functional Data StructuresTyler Prete · LambdaConf 2019
• 19mImmutability MattersRanjeet Singh · Garden City Ruby 2015
• 40mFunctional Programming For The Anxious DeveloperJulian Doherty · RubyConf AU 2017
• 42mTaking inspiration from the functional worldPiotr Solnica · Lambda World 2015
• 35mPersisting Relations Across Time and SpacePiotr Szotkowski · RubyConf 2011