Managing different python versions is hell.
I have seen people sticking with an extremely old but luckily 3.x versions just to avoid it. I don’t like it.
Any suggestions to manage it?
More advanced suggestions than venv and requirements.txt would be appreciated :))
I have seen people sticking with an extremely old but luckily 3.x versions just to avoid it. I don’t like it.
Any suggestions to manage it?
More advanced suggestions than venv and requirements.txt would be appreciated :))
Comments
Or if you’re willing to sacrifice some development speed for consistency, docker-compose with a well-crafted mounts.
Docker could be a good option but we cannot use it.
doing stats subjects ATM — either R, Python or Stata which is a paid sub when I leave uni so why learn it? I did learn python about 15 years ago and it was easy as basic or Pascal to pick up and much better than those primitives. But…
But I also recommend more modern tools like `uv` and `rye` from astral which are faster and written in rust.
Containerisation seems tricky for us due to security concerns by IT office. We will go with a local tooling to manage it.
The Docker engine & CLI are open source, it’s just Docker Desktop that’s paid; but you can use Rancher Desktop instead of that, which is FOSS.
https://code.visualstudio.com/docs/remote/remote-overview
https://github.com/astral-sh/uv
Previously I have used pyenv and it's pretty good but from my short experience with uv I'll say it does more stuff and is on the right track regarding version and dependency management
That does, of course mean you need to learn how containers work... but you know...at least it's a new frustration....
https://www.bitecode.dev/p/relieving-your-python-packaging-pain
https://www.bitecode.dev/p/why-not-tell-people-to-simply-use
I fell back to a manually downloaded Python 3 + Makefile with venv + pip-tools and was simple and good