4 ms·

It’s eating Python too - with Ruff, uv, and an upcoming alternative to Mypy: https://bsky.app/profile/crmarsh.com/post/3lgvhzdfrps26 https://bsky.app/profile/c

• OutOfHere 2y ago
  It's true, but the absence of adoption of these great tools in industry worries me. Most Python developers in industry don't seem to care one iota about handling corner cases of types, and so their code is full of problems. Most Python developers in industry are very untrained, and have no discipline when it comes to types. It makes me sad while working in such a setting.
  • JodieBenitez 2y ago
    I don't know, ruff and uv seem to do great on the adoption. Of course speed is the perceived killer feature, not "types".
    • OutOfHere 2y ago
      Speed and types are different topics. Types are valued by mature programmers who recognize that maintainable and comprehensible projects require aggressive use of types. Short-term thinkers will not value types, and will lead to rapidly decaying and eventually unmaintainable dead projects.
  • anon-3988 2y ago
    I think what will happen is something like numba gets more use. Bad developers can keep writing as they do, but critical infrastructures and functions will be written in natively compiled numba/python code.
    • OutOfHere 2y ago
      I don't think I could agree less. Numba has always had too many errors, bugs, and too inconsistent of a performance gain. Moreover, Python has always been a glue language where numerous packages interact together to form functionality, and this glue usage makes no sense with Numba.
  • gpm 2y ago
    > the absence of adoption of these great tools in industry worries me. UV is a year and a day old. It's not surprising that there isn't huge industry adoption yet. I sort of have the impression that Ruff does have a reasonable degree of industry adoption. Certainly linting does. There mypy alternative doesn't yet exist in a usable state I'm not seeing anything worrying here?
    • OutOfHere 2y ago
      mypy and pyright have been out for so long, and they're in a pretty usable state. Ignoring to use them leads to buggy code that shows its ugly head as unhandled rare conditions begin to manifest themselves. The compounding effects of these unhandled rare conditions ultimately ends the project when the lead subject matter expert exits the team.
  • LordKeren 2y ago
    A large amount of important Python libraries are still missing a good amount of typing - and dealing with third party typesheds is not a good developer experience. The implementation of types in cpython has also been glacially slow and fragmented. To add to the issue, a huge amount of Python in public repositories is untyped and unversioned- which means that LLMs are pretty bad at getting typing correcting in Python, so the debt is growing much faster than the solutions. I honestly think that the most viable solution at this point for typing is a typescript-adjacent solution for Python. I’m kind of surprised Microsoft hasn’t put one out yet