Haja and OrbitDB Code Review

OrbitDB has been in development for some time, a peer-to-peer database for the decentralized web, now got some VC funding from people like Polychain and renamed Haja Networks, out to build on the existing work. Even the name means ‘decentralized’ and they have no offices, seem committed to the cause.

On with the code review, let’s see what they’ve been up to in order to get that funding.

A lot of repositories, all custom, great to see.

orbit-db-store is the underlying abstraction, we will come back to this, for now let’s see what and where our entry point is.

Great documentation. IPFS based data storage, using pubsub and CRDTs. This is looking very sexy.

Modules for event store, feeds, key value, json document stores. This is very comprehensive.

This isn’t rocket science code, it’s just good.

This is sexy, they built their own base library for Conflict-Free Replicated Data Types in Javascript. I’ll be using this library, thanks guys. This is great.

All the different stores implemented, CLI implemented, keeps consistently being great quality code.

There is a lot more code here, even a MVP chat app.

Great abstraction, great encapsulation, just really great code 101 implementations. Mostly done by one guy, haadcode. And been working hard since 2017.

---

Orbit Code Review Conclusion:

This isn’t just an MVP or a prototype, this is a product. This is a very functional, well designed, and well coded distributed data store with support for key value storage, for json document stores and immutable concatenated logs.

I’ll be using this.

I know I shouldn’t be this shocked by an actual functional product. I realized as I was going through this code what I have become. I’m so use to whitepaper only, or a rudimentary proof of concept, that seeing an actual product is almost alien to me.