Faster Python promised with the open sourcing of Pyston project

Python language already runs fast, starting from alternate runtimes to wrapping modules written in C/C++; however, none of these involves the speeding up CPython, the so-called reference implementation of Python, which is perhaps the most widely used version of the language.

Now, an alternate version of the Python runtime, Pyston has released version 2.2 with a very significant new feature: full source code is made available as an open source project under the original Python licensing. Pyston 2.2 offers roughly 30 per cent speed improvement compared to standard CPython, and the software employs just-in-time compilation and other techniques to speed execution.

Albeit, Pyston is significantly different from the other major alternate Python runtime, PyPy, which uses just-in-time compilation to also achieve significant performance improvements.

What the Open sourcing of the Pyston Project means for developers?

The aim of the Pyston project is first to produce a drop-in replacement for the standard Python runtime which can speed up other existing Python deployments without additional effort.

It will make innovation possible in Pyston which can be upstreamed back into Python itself, if the core Python team offers to do so. And the new version of Pyston uses a different approach to achieve this; the base CPython code is altered to help improve performance without breaking the backward compatibility.

The major alternate Python runtime, PyPy remains a large and complex project which has long struggled with being fully compatible with Python, even with Python extensions also written in C. Pyston solves this complexity, starting with changes in the CPython codebase to retain compatibility with it.

How to get Started with the Pyston Project?

The open sourced Pyston v2.2 is available on GitHub, so you can head over there if you are a developer and want to contribute with getting Pyston packaged for additional platforms.

Otherwise, if you want to use Pyston into your projects, it is promised to be as easy as replacing “python” with “pyston” - but if that’s not the case, you can still give Pyston a try and see that it really means to speed up your Python code.