BRIX data architecture | librdx

submited by
Style Pass
2024-12-23 13:30:05

When Swarm.js was launched 10 years ago, nobody did JavaScript CRDT data syncing. These days, there is no shortage in JavaScript syncing frameworks. (No list here, cause I do not keep track anymore!) Some are CRDT, others cache relational data or even run a database in WebAssembly. It is time to level up the game!

The BRIX data architecture implies that “public data” will be like “public WiFi”. You download datasets to your device and you keep them continuously synced. Lots of large datasets. All linked. Why does it make sense?

We already have “supercomputers in our pockets”. 1TB SD card? Easy peasy! With the developments in AI, edge devices finally can make sense of that data. Would a self-driving bus benefit from traffic light timetables, traffic cam data, bus stop cam data? Sure. Are any people waiting on the next stop? Is there a traffic jam? That is all relevant. These decisions better made at the edge. Move it to the cloud and some electromagnetic interference would be able to send a car off the cliff. Or the bus will stop in the tunnel. “Sorry, my humanoid robot is epileptic in elevators!” An average factory floor is one big bunch of electromagnetic interference, by the way.

So far, the human-computer interface was a bottleneck. Now, our devices greatly improve at understanding what we want so they can shovel the data. What we should do here is to change the paradigm from “ingestion” to “syncing”. These are basically two ways to keep the data up-to-date: either ingest into the cloud or sync at the edge. Obviously, in case #1 you are the product.

Leave a Comment