uv: An extremely fast Python package and project manager, written in Rust

Anaconda makes on the order of 100M a year “solving” data science package management. I would argue it has a significantly worse product, attacking a much smaller part of the ecosystem.

It seems easy to imagine Astral following a similar path and making a significant amount of money in the process.

Anaconda isn't free. I don't want to pay per-seat fees for slightly improved versions of open source tools which is why I'm very skeptical of Astral and uv.

One day they're going to tell me I have to pay $10/month per user and add a bunch of features I really don't need just because nobody wants to prioritize the speed of pip.

And most of that fee isn't going to go towards engineers maintaining "pip but faster", it's going to fund a bunch of engineers building new things I probably don't want to use, but once you have a company and paying subscribers, you have to have developers actively doing things to justify the cost.

Enterprises don't care about faster, but they do care an enormous amount about security. Astral is very well positioned here.

In theory, Anaconda solves the next higher level of the Python package management nightmare, namely knowing what versions are compatible with each other. But that could presumably be done on top of uv.

I don't think Anaconda is targeting a smaller part of the ecosystem, rather a different but overlapping ecosystem (remember Anaconda does more than just Python, and there's more to data science than Python). Astral I suspect won't try to move outside of Python, which limits what its tools can be used for.