Ask HN: Why would anyone still use Poetry now that uv exists?

2 pointss-xyz29d ago2 comments

I have been migrating some Python projects from Poetry to uv and the difference in speed is just crazy. I have seen as well several posts in which they do a side by side comparison confirming this.

Not only locally, but as well the deployment of my projects to Cloud Run.

So I am wondering, am I missing something here? Why are people still choosing Poetry for new projects?

2 comments

apothegm29d ago

Concern about proprietary tools. Preferring tools written in the language they’re for. Comfort level. Not feeling they need anything uv offers over poetry. Not bought in to aspects of its architecture (ex: centralized deps). Already having infrastructure and tooling built on poetry. Not perceiving meaningful ROI in learning entirely new tooling for incremental gains.

Poetry is still quite useful. Heck, pipenv is still quite useful. For 95% of use cases, so long as you’re using one of those instead of plain old venv + requirements.txt, you’ll be fine.

kelipso29d ago

I was using pipenv for a while…I think it was useful because I was using different python versions maybe. Then I discovered python -m venv, put all my venvs in a directory in a fast hard drive, along with a venv.sh script that activates the environment with whatever custom stuff I want and I never could find use for any of these environment managers after that. Replication is pyproject.toml and activating the environment is “. venv.sh” what’s hard about that.

j / k navigate · click thread line to collapse