About

I devote about 20% of my time to open-source software maintaining dev containers, docker images and deployment templates for Data Scientists, ML/AI Engineers, and the like.

(GPU accelerated) Multi-arch dev containers¹:

• (CUDA-based) Data Science dev containers
• Mojo dev container

(GPU accelerated) Multi-arch docker images¹:

• (CUDA-based) JupyterLab MAX/Mojo docker stack
• (CUDA-based) JupyterLab Julia docker stack
• (CUDA-based) JupyterLab Python docker stack
• (CUDA-based) JupyterLab R docker stack
  All JupyterLab images include
  • code-server
    → Code - OSS in the browser
  • Neovim
  • Git
  • Git LFS
  • Pandoc
  • Zsh
    GPU accelerated images include
  • CUDA runtime, CUDA math libraries, NCCL and cuDNN
    • including development libraries and headers
  • TensortRT and TensorRT plugin libraries
    • including development libraries and headers
• (CUDA-based) JupyterLab QGIS docker stack
  ---
• (CUDA-based) MAX/Mojo docker stack
• (CUDA-based) Julia docker stack
• (CUDA-based) Python docker stack
• (CUDA-based) R docker stack
  Siblings² of the JupyterLab images without
  • code-server
  • IRKernel/IPython/IJulia
  • JupyterHub/JupyterLab
  • Jupyter Notebook
  • LSP Servers
  • Oh My Zsh
  • Widgets
    GPU accelerated images include
  • CUDA runtime, CUDA math libraries, NCCL and cuDNN
    • without development libraries and headers
  • TensortRT and TensorRT plugin libraries
    • without development libraries and headers

Unofficial and untested releases³:

• Linux/Power (64-bit)
  • code-server
• Linux/RISC-V (64-bit)
  • Pandoc
  • Stack
  • code-server
• Linux/s390x
  • code-server
• Linux/AArch64
  • Neovim

Customised Docker Hub images¹:

• Node.js
  • Node releases: current, active LTS, latest maintenance LTS⁴
• Python
  • Python versions: latest, the last two older⁴
• Julia
  • Julia releases: stable, LTS⁴
• JupyterHub
  • JupyterHub onbuild

Containerised source installations¹:

• Neovim
• Orfeo Toolbox
• QGIS
• Git
• R

Containerised installations¹:

• Git LFS

Deployment templates:

• IDE for Data Scientists: Jupyter
  • JupyterLab + code-server
• DevOps Platform: GitLab CE
• Reverse proxy: Træfik

---

I follow the development of

1. https://gitlab.haskell.org/ghc/ghc
   and
2. https://gitlab.alpinelinux.org/alpine/aports

in order to maintain GHC musl – Unofficial and untested binary distributions of GHC on Alpine Linux.

ℹ️ The multi-arch (linux/amd64, linux/arm64/v8) docker image used to build the statically linked Linux amd64 and arm64 binary releases of

• Pandoc
• Stack
• Juvix

Images available at

• Quay: quay.io/benz0li/ghc-musl
• Docker Hub: docker.io/benz0li/ghc-musl
• GitLab (b-data GmbH): glcr.b-data.ch/ghc/ghc-musl

Linked accounts

• https://gitlab.com/benz0li
• https://gitlab.b-data.ch/benz0li

Footnotes

1. Current os/arch's: linux/amd64, linux/arm64/v8 ↩ ↩² ↩³ ↩⁴ ↩⁵

2. Use cases (not exhaustive): Parent image, dev container, CI pipeline ↩

3. For currently unsupported archs due to missing GitHub runners ↩

4. Base images: Debian (slim): stable, oldstable; Ubuntu: current LTS, former LTS ↩ ↩² ↩³