Today I added some #ArchLinux #AUR packages about #jupyterlite, a light version of #Jupyter that run in WASM on web, as static files service, so reducing security risks. For now, only pyodide (Python scripts) and p5 (js version of #Processing) kernels have be packaged. This took time, due to the number of dependencies, and the incompatibility of kernels with last tinycss python module, and so required a python-tinycss-1.4 AUR package too. You can install directly, using yay, python-jupyterlite-p5-kernel or/and python-jupyterlite-pyodide-kernel, yay will automatically install the package dependencies, but warning that python-tinycss will be downgraded from 1.5.1 to 1.4 (as asked during the process).
The documentation for creating new jupyter instance configuration are here. So on arch (warning, on non-x86_64 architecture (as me, mainly AARCH64 and sometimes RISCV64), you may need to install yay-git, as current stable yay has a bug on compiling packages non-x86_64 architecture.
Installing (as non-root user)
yay -S python-jupyterlite-pyodide-kernel python-jupyterlite-p5-kernel
Creating an instance:
mkdir test
cd test
jupyter lite init
The static web site will be placed in `
jupyter lite build --output-dir /my/loved/static/dir/
If you want to test it quickly, before setting up a web server or uploading it on your favorite host server, you can simply use the included server, by typing:
jupyter lite serve
The instance will then be available at http://localhost:8888/files with both ability to create pyodide and p5 notbooks (if you installed both).
The documentation for creating new jupyter instance configuration are here. So on arch (warning, on non-x86_64 architecture (as me, mainly AARCH64 and sometimes RISCV64), you may need to install yay-git, as current stable yay has a bug on compiling packages non-x86_64 architecture.
Installing (as non-root user)
yay -S python-jupyterlite-pyodide-kernel python-jupyterlite-p5-kernel
Creating an instance:
mkdir test
cd test
jupyter lite init
The static web site will be placed in `
_output` subdir, you can choos another directory by using:jupyter lite build --output-dir /my/loved/static/dir/
If you want to test it quickly, before setting up a web server or uploading it on your favorite host server, you can simply use the included server, by typing:
jupyter lite serve
The instance will then be available at http://localhost:8888/files with both ability to create pyodide and p5 notbooks (if you installed both).
