VSCODESPACE

---

The Goal is to create with vscode devcontainer an universal workspace for datascience.

---

Details about Devcontainers config

• basic_config : config with uv and other cli tools (~2.6GB)
• analyst_devcontainer : with mounted directory and UV manager and Marimo (~2.6GB)
• cuda_devcontainer : with mounted directory, cuda config, UV and Marimo (~5.6GB)

Deployement

1. terminal cli : git clone https://github.com/qwarksky/vscodespace.git
2. Choose your config and inside run : code .
3. vscode => [Ctrl]+[Shift]+P : panel to run commande
4. vscode => Search : Dev Containers: Rebuild and Reopen in Container

Documentation

• devcontainers :
  • https://code.visualstudio.com/docs/devcontainers/containers
• UV :
  • https://docs.astral.sh/uv/guides/integration/docker/
  • https://docs.astral.sh/uv/guides/integration/docker/#available-images
• Marimo notebook :
  • https://docs.marimo.io/guides/deploying/prebuilt_containers/
  • https://docs.marimo.io/guides/deploying/deploying_docker/#prerequisites

CLI :

• listing vscode extensions : code --list-extensions
• listing installed extensions : code --list-extensions | % { "$_" }