All posts
metaopen-source

Prabhava Labs: publishing the decision record with the code

The publishing model behind this studio: every open-source release ships with the reasoning that produced it, including the approaches that were rejected.

An open-source README documents what a tool does. The reasoning that produced the tool, what was tried first, which designs were rejected and why, usually never gets published. For anyone evaluating the code or extending it, that reasoning is often more useful than the feature list, and it is the part that disappears.

Prabhava Labs is a publishing model built to keep that context attached to my own work. Every project on the shelf ships three artifacts:

What the blog covers

Posts here are working notes rather than announcements: performance investigations, API design decisions and the alternatives considered, infrastructure choices for running services cheaply on edge platforms, and post-mortems when a design did not work. When something failed, the write-up says so and states why.

New posts appear in the index, which is RSS friendly, and each repository is public on GitHub as it develops.