Y/stream

A unique attempt to create a multi-master eventual consistent database that syncs many CRDTs (e.g. Yjs documents) efficiently.

• network agnostic
• works everywhere (browser, node, ..)
• small footprint
• documents can have "capabilities" that can be synced with a filesystem.
• syncs many documents efficiently
• Auth* integrated
• (experimental software)

Demos

The demos sync changes to server that I set up on ystream.yjs.dev. You can run your own "server" using npm run ws-server. Note that a server is not strictly required to sync ystreams. It is possible to write p2p connection providers for ystream.

Sync collection to filesystem

• run ./bin/yfs.js --init dir --collection my-collection-name to sync dir to a collection.
• run ./bin/yfs.js --clone clone --collection my-collection-name to clone collection state to a filesystem
• The different directories are now synced. If you open a file in one directory, it will sync the changes to the other directory. This even works on different computers. DONT SHARE SENSITIVE DATA!

Sync collection to a browser application for real-time editing of the files

• cd demo
• Open demo/src/app/ystream.tsx and change the variable collectionName: const collectionName = 'my-collection-name'
• npm run dev
• open [demo]{http://localhost:3000}
• You can run this demo in combination with the file-system demo to see filesystem and editor state synced in real-time.

Work with me

This project is still in the early stages. I'm looking for a company that wants to sponsor this work and integrate it in their product. Contact me kevin.jahns@pm.me.