Hydra is a project to to build a federated/distributed hypergraph database.

It intends to provide a simple API for building applications atop of it which can communicate with each other in near real-time to form a interpersonal shared (hyper)graph with principles of data-sovereignty, anti-siloization, and realtime collaboration at it's core.

It is very early in this project.

Intended features:

• Recording and querying time-series data for a given topic
• Representing and querying event-sourced state for a given topic
• Representing (hyper)edges which define relatinoships between topics (and embeddings)
  • Hybrid hypergraph/vector paradigm
• Declarative + recursive live-querying
• Multi-homed users which are able to roam between servers or self-host without loosing their data or their thought partners

Note that this is very early in development, and it doesn't yet do much of the above, but if you want to see what exists so far you can run the following:

1. Install rust: https://rustup.rs/

2. Install cargo watch (useful for dev workflow)

   cargo install cargo-watch
   

3. Start the server:

   # from the root directory of the repo
   cargo watch -x 'run --bin hydra-server'
   

4. Install wasm-pack https://rustwasm.github.io/wasm-pack/installer/

5. Compile the web client:

   cd web
   cargo watch -s 'wasm-pack build --target web --debug'
   

6. install bun (npm/node might work, but I haven't tested it, and bun is way faster)

   https://bun.sh/docs/installation

   cd examples/webapp
   bun dev