App Mode Launcher - From a Simple Launcher to an Intelligent Workspace Manager

Video Demonstration

A short walkthrough showing the evolution of the project, including the original version, the improvements made during the challenge, and the final application in action.

https://youtu.be/LcBwN6z8cmQ

The Comeback Story

App Mode Launcher originally started as a learning project and a simple application launcher.

After reaching a functional state, development slowed down. The project worked, but it was still far from the vision I had in mind. The codebase was mostly monolithic, customization was limited, and several ideas remained unimplemented.

When I discovered the GitHub Finish-Up-A-Thon Challenge, I saw the perfect opportunity to revisit the project and finally push it further.

During this challenge, I transformed the project from a basic launcher into a much more complete workspace management tool. The codebase was reorganized into a modular architecture, automated tests were added, application detection was implemented, and the overall user experience was significantly improved.

I also improved the project's deployment readiness. The application can now be packaged as a standalone executable, making it easier for users to install, launch, and switch between workspaces without needing to configure a Python environment.

The result is a project that is more maintainable, more scalable, and much closer to its original goal: helping users spend less time setting up their environment and more time focusing on their work.

My Experience with GitHub Copilot

GitHub Copilot played a significant role throughout this project's revival.

One of its most valuable contributions came during the planning phase. Before writing new code, I used Copilot to help create a structured improvement roadmap, breaking the project into smaller, manageable milestones. This made it easier to prioritize features, refactor safely, and track progress throughout the challenge.

Copilot also assisted during the refactoring process, helping reorganize the original monolithic codebase into a more modular architecture. It was particularly useful for generating boilerplate code, suggesting improvements, creating tests, and accelerating repetitive development tasks.

Rather than simply generating code, Copilot acted as a development companion that helped me iterate faster, validate ideas, and maintain momentum while bringing an unfinished project closer to its original vision.

The result was a smoother development process and a significantly improved version of App Mode Launcher delivered within the challenge timeframe.

Thanks for reading!