Server(less) Swift

on YouTube ↗35m

" summary

Claims Swift servers match Java speed, run at ~4x Node.js and ~30–40x Ruby, and use far less memory; examples use Kitura for always-on endpoints and openwhisk (IBM Cloud Functions) for event-driven actions, deployed via Vagrant or cloud with an OpenWhisk CLI. Concrete builds: Weather Gods (three-minute alarm scanning locations and sending push notifications) and BluePic (photo upload + server image recognition + weather enrichment), both wired to Cloudant and App ID; cold-starts under 20 ms claimed but practical lag pushes some work back to an always-on server.

AI-written summary · about

" chapters (14)

1. 0:00Intro
2. 0:21Why Swift
3. 1:55Swagger API
4. 4:39Functions as a Service
5. 10:17Triggers
6. 15:50Weather Gods
7. 17:49BluePick
8. 22:12Server vs Service
9. 24:02Linux
10. 25:59Microservices
11. 28:01Codable
12. 29:35Service and API
13. 31:13Web apps
14. 32:46IBM

" related talks

• 34mServer Side SwiftJens Ravens · App Builders Switzerland 2017
• 41mServer-Side Swift State of the UnionChris Bailey · AltConf 2019
• 30mSwift on the ServerTim Condon · SwiftLeeds 2025
• 25mSwift on Server: Are We There Yet?Gopal Sharma · try! Swift NYC 2018
• 24mServer-Side Swift from ScratchBrandon Williams · .concat() 2018
• 39mWhy Is Swift So Much Better On Server?Gwynne Raskind · UIKonf 2020
• 17mThe Success Story of Server-Side Swift at Cultured CodeVojtech Rylko · ServerSide.swift 2024
• 36mFull stack development with Swift and VaporTim Condon · iOS Conf SG 2020