[Python] What's the best solution to easily use / maintain old legacy codes, specifically...

This python2 software, which hasn't been maintained for 10 years, implements a published statistical model for ribosome profiling experiment data analysis, which is widely accepted and considered the standard in Academia.

https://github.com/ratschlab/RiboDiff

However, it is very hard to install due to dependencies and compatibility issues.

Previously I created a python2 env using conda and successfully installed it. I had to try many times to find a combination of versions of the libraries it depends on that doesn't throw an error. Now I forgot what I did. I exported a YML file of the libraries and versions installed in that environment. Using this YML file, I failed to install RiboDiff.

What I have:
An installed RiboDiff that can run (on Mac).

What I want to achieve:
Brutal force package the runnable RiboDiff that I have and be able to copy it to other computers (Linux) and use it through APIs (like portable executables for many legacy softwares). Or is there a tool to reliably automatically convert these python 2.6 scripts to scripts that can run on python 3.13 with the latest mainstream libraries?

Constraints:
No manually re-implementing. Personal philosophy: I don't want legacy codes to be meaningless. Otherwise, making softwares is a joke: many high level softwares can't outlive 20 years.

What's the best solution?

Continue reading...