Data Stored Everywhere and Nowhere - Building Applications on the Distributed Web
summary
Claim: Dat provides cryptographically addressed, versioned archives for web apps that update over time. Tools shown: Dat CLI, hypercore/hyperdrive/hyperdb, Beaker browser, hyperdiscovery and an ffmpeg pipeline. Demos build a comments-driven website, an offline OpenStreetMap editor, a Science Fair reader and a live streaming feed made of TS chunks; transfers prove fast and peer-accelerating, version history and sparse mode let clients fetch metadata only, discovery uses hashed keys, and integrity relies on Merkle hashes so updates verify correctly.
AI-written summary · about
chapters (18)
- 0:00Intro
- 0:23PeertoPeer
- 2:52Censorship
- 3:47BitTorrent
- 4:22LetsDecentralized
- 4:48Security
- 5:27Feeds
- 11:00Defaults
- 11:56HyperDrive
- 12:12Archive
- 12:44Replicate Archive
- 13:22HyperDiscovery
- 16:07HyperCore
- 17:15HyperDB
- 17:56HyperDB Demo
- 25:02Using Domain Names
- 27:13Metadata Mode
- 27:51In Transit
related talks
43mVersioning, Syncing & Streaming Large Datasets Using DAT + Node
25mManaging changes to your data with dat
23mBuilding and browsing the peer-to-peer web with the Beaker browser
26mSaving Data to the DWeb
22mReclaiming the Web with peer-to-peer protocols
22mDat
25mImagine This: A Web Without Servers
46mThe Datomic Architecture and Data Model