Category
Technical Notes
- · 3 min read
Designing a uv workspace: package boundaries that survive contact with reality
The hardest part of a Python monorepo isn't the tooling — it's deciding where one package ends and another begins. Two rules — change boundaries and interface boundaries — and a worked example.
- · 3 min read
Building an AI Podcast Index: the Project, the Stack, and What You'll Have at the End
An eight-part build-along: a local tool that ingests a YouTube podcast channel, extracts guests and topics with Claude, lets you clip-search by intent, and generates grounded questions for future episodes. Post 1: the demo, the architecture, and one paragraph per stack choice.
- · 3 min read
Why a Python Monorepo: What uv Workspaces Actually Gives You
Most Python monorepo pain in 2024 came from gluing pip-tools, tox, and path hacks together. uv workspaces makes the seam between packages a first-class concept — one lockfile, one resolver, one venv. That's what makes the monorepo worth doing.
- · 3 min read
uv in Docker and CI: lockfile-driven builds that actually cache
Two patterns — multi-stage Docker with uv sync --frozen and GitHub Actions with setup-uv — give you reproducible, fast Python builds with almost no boilerplate.
- · 3 min read
`uv tool` and Single-File Scripts: pipx and Shebang-Python, Replaced
Two uv features outside project management — uv tool for global CLIs and PEP 723 inline-deps for single-file scripts — quietly close out the last reasons to reach for pip or pipx.
- · 3 min read
Migrating to uv from pip + venv + requirements.txt: a pragmatic path
You don't have to commit to a full migration to benefit from uv. The pip compatibility layer is a zero-effort speedup. Move source-of-truth to pyproject.toml only when it earns its keep.
- · 3 min read
Starting a new Python project in 2026 with uv
Six commands, five files, zero `source activate`. A walkthrough from `uv init` to a project that runs tests, ships a CLI, and is ready to dockerize.
- · 3 min read
uv vs pip + venv + pyenv: speed, lockfiles, and the death of `source activate`
Three axes where uv wins and they compound: raw speed from a Rust resolver and a global hardlink cache, a real lockfile instead of a pip freeze snapshot, and a workflow that finally retires source .venv/bin/activate.
- · 3 min read
Why uv Exists: The Pre-uv Python Tooling Mess, and What Astral Consolidated
Before uv, a working Python project needed five tools and a working knowledge of how they fail at the seams. uv collapsed them into one Rust binary because the seams were the problem.
