Programming languages: 'Faster Python' Pyston takes a step forward
The makers of Pyston have released Pyston 2.2, an implementation of Python 3.8.8 that promises to be faster, and they have open-sourced the project.
The Pyston fork of CPython 3.8.8 is available on GitHub with various optimizations that the project claims is "targeted at large real-world applications such as web serving, delivering up to a 30% speedup with no development work required."
Enterprise Software
Key differentiators of Pyston 2.2 (pronounced "piston") are its use of JIT and attribute caching.
"We have tuned and optimized more areas and found additional speedups, particularly in our JIT and attribute cache mechanisms," says Pyston founder Kevin Modzelewski in a blogpost.
SEE: Hiring Kit: Python developer (TechRepublic Premium)
According to Modzelewski's LinkedIn profile, he was a principal engineer at Dropbox for nearly a decade until 2017. Dropbox wrote its backend services and its desktop client almost entirely in Python and it even hired Python creator Guido van Rossum in 2013 to help wrangle its four million lines of code written in Python (Van Rossum left Dropbox in 2018 and is now technical fellow at Microsoft).
Modzelewski notes that Pyston removes "many of the rarely-used debugging features that Python supports" for the sake of speed and to reduce the load that Python debugging creates for computers.
Dropping these debugging features only achieves a 2% performance gain, but given how many computers around the world are running Python code – and the carbon footprint of that code – it is meaningful.
"Of all the computers in the world running Python, 2% of them are executing debugging checks," notes Modzelewski.
"We've disabled those checks and are positioning ourselves as an 'optimized build' similar to binaries without debugging information. Those who still want debugging features can use the "debug build" of stock Python because they are interchangeable."
Modzelewski intends to make Pyston profitable via open source, starting with support services. Anaconda, the maker of a Python distribution focussed on AI and ML development with the NumPy library, runs a similar for-profit operation by releasing an open-source version of its Python distribution and enterprise versions of it.
SEE: Programming languages: JavaScript has most developers but Rust is the fastest growing
Modzelewski also reckons it should be simple for developers to swap out existing Python code for Pyston.
"Working Pyston into your projects should be as easy as replacing "python" with "pyston"," he says.