libsql shill

In the everlasting pursuit of fewer bytes, hops, and cycles, a.k.a my need for speed, I'm exploring every part of the stack in order to identify the largest bits of fat to trim.

I've landed on libsql, the open-contribution SQLite fork, as my go-to for most of my data needs. It's embeddable, lightweight, fast, and very easy to manage and use, and it builds just enough on top of SQLite.

It's relatively barebones, but I get to work with the constraints, which forces me to think about my data, queries, and distribution in a different way. It's a good exercise in minimalism and efficiency.

For most workloads, it's more than enough. If it's not up to snuff, I can always fall back to any capable RDBMS (probably Postgres).

If you’re not part of the solution, you’re part of the Cartesian product.