This is the "Yarn 4 project, but my shell keeps running Yarn 1.22" problem.
You'll see an error like:
This project's package.json defines "packageManager": "[email protected]". However the current global version of Yarn is 1.22.22. ...the project is meant to be used with Corepack...
This gist is the "do it once, stop suffering" setup for macOS (Apple Silicon) where:
python3 is too new for your project (ex: Python 3.14 breaks/doesn't have stable wheels for stuff like Pygame yet)/usr/bin/python3 (often 3.9.x) and it silently wins when Homebrew isn't configured how you thinkpip installs going into the venv (no PEP 668 drama)This documents the exact failures I hit and the final working config.
| #!/usr/bin/env python3 | |
| """ | |
| Batch-convert .mov/.MOV QuickTime videos in the current directory: | |
| 1. Find all .mov/.MOV files. | |
| 2. For each: | |
| - Convert to MP4 (H.264, no audio, slightly reduced quality). | |
| - Generate a high-quality GIF from the MP4. | |
| - Extract JPEG screenshots from the GIF every 1 second. | |
| - If everything succeeds, delete the original .mov/.MOV file. |
This is a practical, opinionated guide to Python virtual environments on POSIX/macOS/Linux with a few Windows notes. It shows the cleanest workflows that don't bite you later: create/activate, install via pip and requirements.txt, run scripts with reliable logging/printing, deactivate/uninstall/remove, common gotchas, when to upgrade pip/setuptools/wheel, and a Zsh prompt that shows the active venv in red. It also explains what a venv does and doesn't do (spoiler: you can still build Go, run Docker, etc.).
Python versions and packages vary per project. A venv creates an isolated interpreter + site-packages so each project can pin its own dependencies without polluting system or Homebrew installs. On macOS/Homebrew you usually only have python3 globally; once you activate a venv, you get a project-local python.
A small Python CLI that transcribes audio using Vosk.
If you pass an .mp3, it automatically converts to a 16 kHz mono WAV via ffmpeg and then transcribes. If you pass a .wav, it transcribes directly.
.mp3 or .wav.txt with the same base name as the input (e.g., audio.mp3 → audio.txt)
A small Bash function that prints only the meaningful parts of your Git diff, skipping asset/vendor directories and binary-ish files. Output is grouped by file with a clear header, making it ideal for quickly reviewing changes or pasting compact, high-signal code context into LLM prompts.
node_modules/, .yarn, .astro, and common binary/extensions by default.
| #!/usr/bin/env python3 | |
| # Tetris clone in Python + Pygame | |
| # No assets needed. Run: python tetris.py | |
| # Controls: | |
| # Left/Right: move | |
| # Down: soft drop | |
| # Up / X: rotate CW | |
| # Z: rotate CCW | |
| # Space: hard drop | |
| # P: pause |
A robust, cross-platform shell function for quickly sharing a directory’s structure and previewing file contents.
Ideal for collaborating with others, sharing context with ChatGPT or other LLMs, or just auditing a codebase without manually sifting through files.
| #!/usr/bin/env python3 | |
| import argparse | |
| import os | |
| import sys | |
| import re | |
| # directories to skip entirely | |
| EXCLUDE_DIRS = {'node_modules', 'target', 'bin', '__pycache__', 'build', 'dist'} | |
| MARKDOWN_LINK = re.compile(r'(!?\[.*?\])\((.*?)\)') # matches [text](url) and  |