Rebooting the Blog with Pixi and Hugo
Migrating from Jekyll to Hugo with Pixi for a faster, more reproducible blog setup.

This is my personal blog. I write about technology, open science and open data. I am involved in a number of open source projects, the biggest are xtensor and xsimd. I also work a lot on technologies in the Project Jupyter space. I did my master studies in Robotics, Systems and Controls and still use ROS on a daily basis.
I am currently employed at a small & very fine software consultancy in Paris, called QuantStack where we are lucky to work on mostly open source software.
Migrating from Jekyll to Hugo with Pixi for a faster, more reproducible blog setup.
boa is a new conda-build alternative that produces conda packages -- and also aims to enable a "source distribution" and easy re-configurability of conda packages
I have worked on finishing drag'n'drop in ipywidgets, and adding dynamic loading to JuppyterLab extensions.
CAD programs have a constraint solver for sketches since forever, but they are generally missing from graphics software.
I have been curious to figure out where the Siri knowledge queries go, and what is returned.
It's fundamentally unfair to create Open Data without working on free tools to derive useful knowledge from this data.
A small trick to demonstrate how to achieve expanding or wide source code with Jekyll
Conda is often thought as a Python package manager, but it works great for C++, too! Here is how:
Just a quick hello on the new blog.