Unlocking API Potential: Why Model First Wins
summary
Argues that a model-first registry beats schema registries by letting tooling generate APIs from domain models, shows Hasura introspecting a Postgres database and using track all to auto-generate relationships and a production GraphQL endpoint, cites Hasura adoption and OSS numbers, and proposes a declarative Graph Data Specification to declare what fields are selectable/aggregatable so permissions, caching and SQL execution are inferred rather than hand-coded.
AI-written summary · about
chapters (10)
- 0:00Intro
- 2:19Who is Tanmai
- 3:38What is Hasura
- 6:19Defining Code First, Schema First, Model First
- 10:00Model First Explained
- 10:23Code First Approach
- 14:36Schema First Overview
- 23:44Upcoming Changes in Hasura
- 25:25Spicy Takes on GraphQL
- 26:25Thank You
related talks
22mUnlocking API Potential with Hasura's GraphQL Engine
32mModernize development and activate legacy data with GraphQL
27mAn API First Approach to Adopting a Distributed SQL Database
23mOpen Source, Graphs and GraphQL: The next frontier of modern applications
23mBuilding APIs you want to hug with GraphQL
18mEmpowering Kubernetes Workloads with GraphQL
40mOpening Keynote
25mGraphQL Data Dictionary