Hacker News Reader

Yup, I ended up learning docker because Python packaging was such a dumpster fire.

I'll never forget the day I discovered that pip didn't do any dependency resolution. You had one job, and yet...

Until uv, I used docker to run python projects locally. Got so fed up with venvs and versions and slowness and lack of wheels and spending so much time mucking around with it when cloning or updating a repo.

Uv hasn't fixed all issues, but made it far less painful.

apt install …

When people all have these special needs they never show their code.

So apt install worked if you had a decent Linux box and the correct versions of C and Fortran libraries were available.

I worked with data, so tended to use conda and still have loads of problems. Pre compiled wheels and uv have made the situation a lot better.

When you do apt install, the maintainer made sure that the correct versions of whatever dependency was indeed available. That's the entire point of a distribution.

Yes. But many of the dependencies I needed were either not packaged or so old that I couldn't use them.

Seriously, data stuff was a massive massive struggle with Python for ages, and uv and compiled wheels have made it much better.

Reply on news.ycombinator.com