4206be2593
- Add `--rm` to auto-remove containers after exit - Add `-u $(id -u):$(id -g)` to run as host user (fixes PermissionError on Linux where container's UID 999 doesn't match host user) - Add `-e HOME=/tmp` so Typst can create its package cache when running as an arbitrary UID (without this, `rendercv render` fails) - Update both user guide and developer guide docs
2.0 KiB
toc_depth
| toc_depth |
|---|
| 3 |
Dockerfile
What is Docker?
Docker lets software carry its entire working environment with it: the right language runtime, libraries, and configuration, all bundled into a single file called an image. Think of an image as a frozen filesystem where everything is already installed and configured correctly.
When you run an image, Docker creates a container: a live, isolated instance of that environment running on your machine. When you're done, you can delete it without a trace. Your actual system stays untouched.
Why Docker for RenderCV?
RenderCV installs easily with pip install rendercv if you have Python. Most users don't need Docker.
But Docker makes sense if you want:
- No installation at all — no Python, no packages, nothing added to your system
- A reproducible environment — the exact same setup on every machine, every time
- To bypass restrictions — some systems block software installation but allow containers
The RenderCV Docker image is a ready-made environment with Python and RenderCV pre-installed. Just run:
docker run --rm -v "$PWD":/work -u $(id -u):$(id -g) -e HOME=/tmp -w /work ghcr.io/rendercv/rendercv new "Your Name"
How the Image Gets Published
Docker images are stored in registries, which are servers that host images so anyone can download and run them. Docker Hub is the most popular, but GitHub has its own called GitHub Container Registry (GHCR).
When RenderCV publishes a GitHub release, the release.yaml workflow automatically builds and publishes the RenderCV image to GHCR at ghcr.io/rendercv/rendercv.
When users run docker run ghcr.io/rendercv/rendercv, Docker automatically pulls the image from the registry if it's not already on their machine.
Learn More
To learn more about writing Dockerfile, see the uv's guide on Docker.