“Don’t reinvent the wheel.”

It is the first piece of advice every developer gets. It is efficient, practical, and safe. It is also the reason why so many of us have no idea how our tools actually work.

We live in the era of the Black Box. We npm install complex logic, we send HTTP requests without knowing how a TCP handshake actually negotiates. We are becoming great at gluing things together, but we are forgetting how to build the glue.

I created mendoza.gg to write along the lessons I learn on this journey.

The Skeleton Philosophy

I am not here to build production-ready software. I am here to build Skeleton Projects.

A Skeleton Project is a stripped-down, raw implementation of a complex piece of technology. It is the “bones” of the software, devoid of the flesh, skin, and safety features that make modern tools user-friendly and probably production ready.

If I want to understand how a Game Engine works, I won’t follow a Unity tutorial. I will write a renderer, a loop, and a physics update function from scratch. If I want to understand interpreters, I will write a parser that can do basic math, even if it crashes on a syntax error.

The Rules of Engagement

For every Skeleton Project I tackle on this blog, I will try to follow these rules:

  1. First Principles: If I don’t understand a concept, I don’t implement it until I can write it down on paper.
  2. Destruction is Success: Breaking the code, hitting segfaults, and creating memory leaks is part of the process. I learn more from a crash than a successful compile.
  3. Rule 3:

What’s Coming

I am currently lining up the first few victims for the chopping block:

  • Skeleton Engine: A 2D game engine built without a pre-made engine.
  • Skeleton Interpreter: A language parser that can (barely) run code.
  • Skeleton HTTP: A web server that speaks the raw text of the internet.

This blog is my lab note collection. It won’t always be pretty, and the code definitely won’t be production-ready. But it will be honest.

No magic. Just bones.