These posts go into more detail about particular commits to the Midestro project. Mainly why I thought I needed to make them and how I came to the implementation.

• The Beginning (October 23, 2024) - TL;DR - Made a page that wasn't very good at being a game yet. Took notes about what to do better for the next several iterations.
• Iterating Towards Playability (October 24, 2024) - TL;DR - Added more debugging and worked on UI / UX concerns with selecting notes to play and seeing them at the right places on the canvas.
• Adding Interactivity (October 25, 2024) - TL;DR - Added some game options and better visual feedback for scoring points during the game.
• Getting to Feature Complete (October 26, 2024) - TL;DR - Added a scoreboard, fixed bugs, and got to Feature Complete!
• Modest Optimizations (October 28, 2024) - TL;DR - Trimmed down on loops and iterations in loops while the game interval ran.
• Effective Optimizations (October 29, 2024) - TL;DR - Added Precached rendered notes and measured performance with browser developer tools.
• Unit Testing and Software Development with Llama 3.3 70b (November 23, 2024) - TL;DR - Wrote Unit Tests and rewrote complex method that had bugs in it.
• Creating 3D Version with Agents (March 19, 2026) - TL;DR - Used Claude Code and OpenCode to update Midiestro alot.

Local LLM Experiments

These are experiments I come up with to see how different LLMs handle similar tasks. The exact process of these experiments is still evolving

• Blog Post Styling with LLMs - TL;DR - I spent hours asking several LLMs to make my Blog Posts look nicer and then asking them to compare their changes to each other to determine which LLM made the best Patch.

Homelab Experiments

Container experiments run on a laptop to learn about infrastructure, databases, messaging, observability, and more.

• 39 Container Experiments on a Laptop (May 3, 2026) - TL;DR - Ran 39 docker-compose experiments across 8 domains on a modest laptop and learned more from failures than successes.