b-data / data-science-devcontainers

(GPU accelerated) Multi-arch (linux/amd64, linux/arm64/v8) Data Science dev containers for R, Python, Julia and Mojo
MIT License
14 stars 9 forks source link

Basic usage questions #14

Open jpvelez opened 4 days ago

jpvelez commented 4 days ago

Hello! This project looks super cool, and I'm trying to work out how to use it.

My ultimate goal is to use glcr.b-data.ch/jupyterlab/r/geospatial as a development environment in vscode (the desktop version, not the browser version).

A few questions!

jpvelez commented 4 days ago

Okay, I did a lot of reading about devcontainers and I think: 1. devcontainers are awesome and 2. I answered the first 3/4 questions above.

So I still have the last question, about persisting the workspace.

And new question, if I use this devcontainer what's the intended mechanism for adding my own R and Python packages? Should I make my own Dockerfile based on yours and point the r-geosptial devcontainer at that one instead?

benz0li commented 3 days ago

Is it possible to override the directory where files are mounted to something other than /home/vscode? (Will that screw something up?)

The idea is that you store your work within the (default) user's home directory (/home/vscode, e.g. folder ~/projects) – as you would with a regular Linux distribution.
ℹ️ Persisting the home directory also ensures that dot files/folders (e.g. settings) are persisted.

By default, mount type volume is used, so Dev Containers behave the same as Codespaces.
ℹ️ See also [Data Science dev containers] Do they work with bind mounts?

Also, why would workspaces be persisted, and where?

/workspaces is persisted by default. Where?