My Experience with GitHub Copilot

Duke recently started using GitHub Copilot and integrated it with Python and .NET languages, testing it on various IDEs including VSCode, Visual Studio, Rider, and PyCharm. Setting up GitHub Copilot wasn't too difficult, but users of Visual Studio need to be on version 17.10 or later. GitHub Copilot is compatible with major IDEs thanks to various extensions, making it versatile and easy to use across different setups. The code generation quality of GitHub Copilot is impressive, especially in following coding conventions, such as PEP 8 standards for Python. To get the best results from prompt-based code generation, it's crucial to be specific with the prompts. GitHub Copilot also provides function name suggestions, but clear and descriptive names are necessary for better suggestions. One of the standout features of GitHub Copilot is its ability to explain code, providing contextual understanding of single lines and entire sections. However, the auto-fixing code feature is lacking in some areas, such as correcting errors in algorithms. Despite its imperfections, GitHub Copilot is a solid tool for code generation and explanations, making it worth trying for coders, especially those working with Python or .NET. Overall, GitHub Copilot has been a valuable addition to Duke's coding toolkit.