bang. bang. bang. https://gun.eco
censored on twitter, help me break it?
https://twitter.com/marknadal 

@marknadal’s gunDB really fast and I always wondered if I could build a decentralized data lake for data science. I think it would be powerful thanks to the circular references, S3 compatibility, and its real-time capabilities (it’s so cool!). Do you think it would think a good idea?

Currently researching tools and software to make real-time, fast & snappy crypto apps.

I’m even thinking of WASM plugin API to pre-process the data as it is sent in the database !

Insofar as you need large batches of data to fit a statistical models, or even AI models, I thought we could use GunDB to create a data lake that could easily be fetched from everywhere, quickly, etc…

S3 API compatibility (not vendor locked to S3, just to be clear for others reading, S3 API is a shared standard across a lot of even open source projects these days).

What were you ideas for the data lake? That sound cool. Curious what the lake part is.

BDFL of Quilibrium, Eng @ Farcaster, ex-Coinbase, always opinionated, never hydrated

/quilibrium

Oooh, it’s cool! It’s kinda what I wanted to make, but my version wouldn’t be a service but more a library dependent on the user’s own infrastructure for the database. You provide the peers, and I provide the tech.

We’re about to launch something like that but fully E2EE, though it’s more than S3 compatible: https://www.quilibrium.com