Bootstrapping Your Local Python Environment
summary
Clear rules: no sudo, use explicit per-project interpreters, and prefer repeatable installs. Demonstrations show installing multiple Python versions with pyenv, switching globals and locals (examples: 3.10.3 and 3.9.11), creating isolated virtualenvs and installing packages there, and using pipx for command-line tools. Dependency workflows use piptools to compile locked requirements and Docker to freeze builds. The live demo switches versions and sandboxes successfully, with only minor demo friction.
AI-written summary · about
chapters (10)
- 0:00Intro
- 0:24LET'S SET OUR INTENTION
- 2:41PYTHON LIVES IN MANY PLACES
- 4:26STAY ZEN
- 6:44DO NOT USE THE SYSTEM PYTHON
- 7:43PYENV Change the global Python version on a per user basis Per project Python versions Override the Python version with an environment variable Search commands from multiple versions of Python at a time
- 18:52VIRTUAL ENVIRONMENTS ARE INCLUDED IF YOU HAVE SIMPLE NEEDS, THIS IS FOR YOU!
- 22:19WHAT ABOUT THE "USER SCHEME"?
- 23:41BUT WHAT ABOUT... Anaconda Python and Conda ActiveState Python
- 25:27BUT I WANT REPEATABILITY AND SIMPLICITY... Simple is better the complex
related talks
38mBootstrapping your Local Python Environment
30mPython Versions and Dependencies Made Easy
24m"Invisible Walls: Isolating Your Python"
37mMaking Sense of Python Virtual Environment Tools and Workflows
30mHow to give your Python code to someone else
29mDon't Cross The Streams - An Introduction to Virtual Environments
40mA Few Python Tips
28mBreaking Free from Virtual Environments Python's New Paradigm for 2026