The Meaning of LIVE
summary
Argues the current cloud/stack is a viral, inhumane architecture and proposes data-first software: substrates that are a persistent code+data store with direct-manipulation UI, live programming, and HTTP/JSON as the foreign-function interface. Points to Smalltalk and HyperCard as precedents, notes hot reloading and playgrounds as partial wins, and targets the Gap between spreadsheets and full-stack apps (thousands to low millions of rows). Calls for work on schema-change handling, end-user version control, supply-chain (source-only) trust, modularity, and alternative UI metaphors.
AI-written summary · about
chapters (25)
- 0:00Introduction
- 3:57Stack vs substrates in programming
- 7:28Stack defeated substrates historically
- 10:40Data first software concept
- 13:07Trustworthy source and supply chain
- 14:11Research agenda for substrates
- 17:26Need unified field and naming
- 19:36Platform shifts and future turbulence
- 21:03Q&A
- 21:24Disintermediate data from code
- 22:57Database sufficiency and persistence
- 24:03Scope of small data
- 25:29Spreadsheets as programming
- 27:11Scientist user needs
- 28:58Modularity and business models
- 31:51Legibility and interoperability
- 33:56JSON HTTP as new FFI
- 35:00Substrate ecosystems
- 37:08Beyond grid with cross linked trees
- 38:3410 years of LIVE
- 40:23Visual programming limits
- 41:06Sweet spot for substrates
- 42:31Universality vs specialization
- 44:21Insularity vs interoperability
- 46:08Closing remarks
related talks
35mKeynote: Fireside Chat on LiveView, (...)
21mModifiable Software Systems: Smalltalk and HyperCard
20mBuilding a cloud-friendly application
57m🚨 Live On the Edge 🚨
01h 13mKeynote: Phoenix LiveView - Interactive Apps without Javascript
22mBuilding WebApps Like It's 1972 🧙♂️by Leandro Ostera
32mReactive Data on the Full Stack
51mSurchargés Live: Episode 2!