Why Writing Blockchain Applications is Hard, and How FP Can Help
summary
Claim: a small domain-specific language compiles abstract specs down with cascading proofs so a 20–50 line spec avoids the 2,000 lines of hand‑written code the demo shows (contract ~300–400 lines, client ~400, plus JS glue). Tooling claims: a Racket‑based DSL, endpoint projection, game semantics and category‑theory semantics, targetting Solidity and JavaScript now with EVM and SNARKs planned. Demo attempts a rock-paper-scissors app via Electress and intermittently fails but reveals the compilation and verification workflow.
AI-written summary · about
chapters (14)
- 0:00Intro
- 7:38Challenges
- 17:30General Purpose Language
- 17:56Multiple levels of abstraction
- 19:04Contract language
- 20:13Virtual machine
- 24:02Cascading verification
- 26:56What do we want to automatically verify
- 29:39How to keep complexity in check
- 31:15Category sharing
- 32:55The blockchain agnostic model
- 35:31Smart contracts
- 40:23Mathematical essence of blockchain
- 41:20Conclusion
related talks
04mInterview
45mFunctional Smart Contracts
43mThe Big Problems Facing Functional Programming & a Possible Solution
54mQ & A Session With Functional Conf Speakers
45mDeveloping DSLs with FP: Exploring 3 Strategies Through Examples
36mConcrete Functional Programming
39mFunctional Programming And the Real World
58mKeynote: Why Functional Programming Matters