Brave New World, tales of PureScript and Haskell in production
summary
A payments team moves from typelevel Scala to a stack that starts with PureScript -> Haskell migration, using Nix mono-repo, Terraform and AWS Lambda serverless for deployment, then drops PureScript and serverless after heavy wrapper maintenance, slow parallel CI and cold starts. One app rewrites in a week, production Haskell services total about 10,000 lines of Haskell, give fast cold-start-free performance, mature libraries, and shorter ramp-up (3–4 weeks) versus months for purely functional Scala.
AI-written summary · about
chapters (7)
- 0:00Start
- 0:05Brave New World
- 1:36Write Haskell for Snoop Dogg
- 15:18Purely Functional and Serverless
- 20:38Novelty Budget
- 22:17Serverless considered harmful
- 27:59What are you optimizing for?
related talks
36mFunctional Programming in Serverless World
32mHeadfirst into Haskell
41mEnd-to-end Purescript with Erlang/OTP
31mWhy Haskell is a Terrible Choice for Startups (And Why We Picked It Anyway)
57mReal-world Haskell: Good, Bad, & Ugly
54mQ & A Session With Functional Conf Speakers
52mBuilding Scalable Web Applications with Haskell
01h 02mFrom Monolith to Serverless
