We're building the fast, deterministic and reproducible package manager Pixi for scientific and robotics workflows. Based on @condaforge & @robostack packages.
We just made it really easy to switch from conda to pixi. You can now use the pixi init - - import environment.yml to quickly port a conda project. You automatically get a pixi.toml, lockfiles and a powerful task system.
Fast & reproducible @nvidia CUDA development with Pixi is easier than ever, thanks to the new capabilities of `pixi-build`. In our blogpost we build a simple kernel to run Mandelbrot on the GPU!
Learn more:
More and more multi-output recipes are now building through conda-forge's new rattler-build based staging.
Fast, modern, Rust-powered builds are landing in conda-forge, one recipe at a time.
Security tip for conda package authors: prefer environment-variable JSON files over activation scripts.
Activation scripts run arbitrary shell code on every activation. Plain env-var JSON does the same job in most cases, with none of the code-execution risk.