Implementing A User-Centric Datastore
summary
Claims a user-centric personal cloudlet that gives consumers object-level permissions and logs access while developers use a reusable type system; implemented as an OpenIE/PEAT microservices platform (node.js, Mongo/Couchbase, Swagger, message queues) using JSON Web Tokens to bind app+user sessions. Permissions live per object, types are immutable types and published to a registry, a fertility-tracker integration demonstrates end‑to‑end flows but the live demo and analytics engine are unreliable.
AI-written summary · about
chapters (28)
- 0:00Introduction
- 1:11OpenEye
- 2:25OpenID Platform
- 5:10Cloudless Framework
- 5:40Objectives
- 7:10Challenges
- 7:33Implementation
- 8:17Components
- 9:55Data Access
- 10:45Aggregator
- 10:59JSON Web Tokens
- 11:32Session Tokens
- 12:05Cloudless
- 13:18Type Builder
- 14:46Permissions
- 15:25User Dashboard
- 15:51Use Case
- 18:15Data Types
- 18:48Permissions Dialogue
- 20:17Permissions Dashboard
- 20:27Analytics Engine
- 21:10Live Demo
- 22:13Login Process
- 25:50Demo Area
- 30:21Trust
- 30:56Offline Access
- 32:08Type Duplication
- 34:13Changing Data Types
related talks
51mImmutable data stores and CORS for safety, flexibility and profit
13mArchitecting a Data Privacy Platform to Scale to Billions of Identities
10mPrivacy: The Next Frontier
47mThe future of Real-time/offline/data
25mOffline-first data: Getting Bigger
01h 28mData privacy vault 101
34mData Storage: NoSQL Toasters and a Cloud of Kitchen Sinks
29mData Stored Everywhere and Nowhere - Building Applications on the Distributed Web