Browse Teach Curate

Reproducible Python Environments

A result is reproducible only if the interpreter version and the full transitively-resolved dependency set are pinned; lockfiles (uv/pip-tools/conda-lock) record exact versions and hashes so an environment can be rebuilt identically later or on another machine. 'Works on my machine' is almost always an unpinned transitive dependency.

jeremy Video

Isolating Project Dependencies with venv Virtual Environments in Python

A virtual environment is a self-contained directory holding an independent interpreter reference and package installation tree, giving each project its own dependency set isolated from the global ins…