Sign inSign up

chungc/divenb

By chungc

•Updated 1 day ago

JupyterLab + VS Code for DIVE. Bundles MIT Scheme, ROOT, TeX, R, Chrome & Hermes Agent. Multiarch.

Buildkit cache
Image
Developer tools
0

1.1K

chungc/divenb repository overview

⁠DIVE Notebook Image (divenb)

divenb is the general-purpose Jupyter notebook image for DIVE, the AI-enhanced workspace of Project Maieutic⁠ (specification-driven and agentic computing education at CityUHK). One image, no setup: JupyterLab and VS Code in the browser, a scientific computing stack (MIT Scheme, ROOT, TeX/LaTeX, R, Octave, Cling/C++), plus Chrome, Obsidian, and the self-learning Hermes Agent you can inspect and adapt.

⁠Get the image

The image is multiarch (linux/amd64 and linux/arm64); Docker pulls the matching architecture automatically:

docker pull chungc/divenb:0.2.86
docker run -it --rm -p 8888:8888 -p 8000:8000 chungc/divenb:0.2.86

Ports: 8888 JupyterLab, 8000 code-server (VS Code in the browser).

To pin an architecture explicitly, use the suffixed tags 0.2.86-linux-amd64 or 0.2.86-linux-arm64. Verify the multiarch manifest with:

docker buildx imagetools inspect chungc/divenb:0.2.86

⁠Use it for your course

The image has nbgitpuller installed, so course notebooks are opened by clicking a pull link (no git setup needed). With the image running locally on the default port, click to pull a course into JupyterLab:

(Use your port instead of 8888 if you map a different one.) On a JupyterHub, the same links take the form <hub>/hub/git-pull?url=<repo>&branch=<branch>; the nbgitpuller link generator⁠ creates them for any repository, branch, or file to open.

⁠Versions

Tags follow the DIVE release scheme (semver). See the tag list⁠ for all published versions. The cache1 tag holds the shared build cache; keep it when rebuilding.

⁠Building the image

The image is built from the jupyter repository⁠ (divenb/Dockerfile). The repository README⁠ documents the build targets and the multiarch workflow (single-machine cross-build, per-arch native builds on Linux/Apple Silicon, and stitching the pushed halves into one multiarch tag) in detail.

Course notebook images: chungc/cs1302nb⁠, chungc/cs2310nb⁠, chungc/cs5483nb⁠

Tag summary

Content type

Image

Digest

sha256:1322d1d79…

Size

9.4 GB

Last updated

1 day ago

docker pull chungc/divenb:0.2.86