3 ms·

I see the potential to speed up code a lot. Say that you could translate Python to C++ or Nim that would speed up the program a lot. You could profile and trans

by acd 6y ago

I see the potential to speed up code a lot. Say that you could translate Python to C++ or Nim that would speed up the program a lot. You could profile and translate select hot often used code paths. This has the potential too save a lot of energy by using more efficient programs.

dunefox 6y ago
That seems like a lot of effort when Julia is both high-level and efficient already.
- RMPR 6y ago
  Let's rewrite everything in Julia.
  - dunefox 6y ago
    There's no need to rewrite anything: https://github.com/JuliaPy/PyCall.jl https://github.com/JuliaPy/PyCall.jl But yeah, I won't start any new projects in Python.
- gnulinux 6y ago
  Ok but I have 3 million lines of Python code, so that doesn't help me. I'm not gonna spend a year or two writing everything in Julia.
  - klyrs 6y ago
    I hear that you can use Python from Julia and vice versa. You could technically get started immediately and strangle the python away.
  - dunefox 6y ago
    Good thing you don't have to: https://github.com/JuliaPy/PyCall.jl https://github.com/JuliaPy/PyCall.jl
klyrs 6y ago
Cython already translates Python to C. And if the translation introduces a bug, the Cython developers can find a solution. This ML crap is inexplicable, and nearly guaranteed to introduce subtle bugs and performance regressions.