Why VS Code for C# Developers

If you’re a C# developer who has relied on a full Visual Studio IDE for years, Visual Studio Code (VS Code) might look like too small a tool for serious work. In reality with big solutions you will find full Visual Studio more productive but Visual Studio Code now has some specific features that makes it more useful especially for small project.

C# Developer Toolkit

For productive C# development in VS Code you only need a few well-chosen pieces: the .NET SDK and dotnet CLI, the official C# extension (OmniSharp) for IntelliSense and debugging, a Test Explorer adapter for running unit tests, and a couple of linters/analysers (EditorConfig, Roslyn analyzers or StyleCop) to keep code consistent. These core tools give you a fast edit-build-test loop while keeping your environment simple and reproducible.

The extension C# Dev Kit from Microsoft bundles all these tools in a single install, making it easy to get started with C# development in VS Code. The problem is that it does not rival with all the feature of full Visual Studio IDE so the real question is:

Why should a C# developer use VS Code instead of full Visual Studio IDE?

GitHub Copilot

The experience of GitHub Copilot in VS Code is +really better that what you got on other IDEs*. Not only I use VS code insiders and I can work with the pre-release version of Copilot but the whole experience is more fluid and integrated.

Clearly you can work with both Visual Studio and VS Code opened in the same folder, you can just use the tools more useful for the specific task, (as an example Debugging in Visual Studio is WAY MORE powerful that VS code) so you can benefit of both worlds.

Codespaces

I usually have side projects that I develop on my workstation, then on my laptop etc, and one of the most annoying things is to keep all the tooling in sync. Since I often works with MongoDB or Elasticsearch, it is convenient configuring a devcontainer environment that will declare dependencies from other containers. This is an example of docker compose file for my devcontainer.

version: '3.8'

services:
  app:
    image: mcr.microsoft.com/devcontainers/dotnet:1-10.0
    volumes:
      - ".:/workspace:cached"
    command: sleep infinity
    depends_on:
      - mongodb
      - elasticsearch
    networks:
      - app-network

  mongodb:
    image: mongo:6.0
    container_name: devcontainer_mongodb
    restart: unless-stopped
    ports:
      - 27017:27017
    volumes:
      - "mongodb_:/data/db"
    networks:
      - app-network

  elasticsearch:
    image: docker.elastic.co/elasticsearch/elasticsearch:9.2.2
    restart: unless-stopped
    ports:
      - "9200:9200"
    environment:
      - discovery.type=single-node
      - xpack.security.enabled=false
      - ES_JAVA_OPTS=-Xms1g -Xmx1g
    volumes:
      - "es_:/usr/share/elasticsearch/data"
    networks:
      - app-network

volumes:
  mongodb_:
  es_:

networks:
  app-network:
    driver: bridge

Thanks to this simple configuration I can have a consistent development environment everywhere, and since I can connect to codespaces from a local instance of Visual Studio Code, I get the best of both worlds. Since VS Code will connect to the remote environment I can simply look inside MongoDB or Elasticsearch with the proper extensions installed in my local VS Code instance, while the code is running in a remote environment.

Thanks to configuration I can specify envirnonment variables, ports to expose, and all the dependencies I need for my project. This is a section of my .devcontainer file

	"containerEnv": {
		"TEST_MONGO_INSTANCE": "mongodb://mongodb/test",
		"AZURE_MODEL": "gpt-5-nano",
		"AZURE_EMBEDDING_MODEL": "text-embedding-3-small",
		"ELASTIC_TEST_URL": "http://elasticsearch:9200"
	}

I usually install also the extension for Claude Code that works perfectly in my environment.

Using devcontainer with codespaces is not the only option, if you have a powerful workstation you can use docker desktop to bring your container to life locally, the experience will be the same, but if you need to work from different machine, you just need docker and the rest is automatic.

Conclusion

If you are a C# developer that wants to leverage the power of GitHub Copilot and Codespaces, Visual Studio Code is the natural choice. The experience with Copilot is simply better. For some specific tasks you can always rely on full Visual Studio IDE, but for the rest VS Code is lightweight, fast, and with the proper configuration you can have a consistent development environment everywhere.