The Python compiler Nuitka, written in Python, is said to deliver performance benefits in version 1.0 through improved memory management.
The team around Nuitka developer Kay Hayen has completed version 1.0 of the Python compiler. Numerous bug fixes have been incorporated into the latest release. According to the announcement, the focus was on improvements that should contribute to higher scalability and performance – for example through optimized memory management.
Reduced compilation times within Nuitka and lower memory consumption should give the Python compiler more speed, especially for larger compilation tasks. According to the blog post, one of the ways the team achieved scalability was by avoiding initial variable merge traces that occurred when merging a variable that was only used in one branch.
Clean-up work for more comfort and speed
In addition, the Nuitka developers have devoted a lot of time to plugins and YAML configuration. For example, cleaning up some DLL configurations should make it easier to use. However, the documentation for the YAML configuration is still considered incomplete. Among other things, there is a lack of examples that enable the community to add missing dependencies, data files, DLLs or even anti-bloat patches. This is why, for example, the option tailored to YAML files provided by users was also added --user-package-configuration-file new added. With this option, Nuitka should be able to pass a package configuration that allows the addition of DLLs or files.
The Python compiler Nuitka, written in Python, is compatible with most current Python versions (2.6 and 2.7, 3.3 to 3.10) and can be used on almost all platforms on which Python is also used. Nuitka creates executable files that do not require a separate installer. Additional data can either be integrated directly or attached as files. As a replacement or in addition to the Python interpreter, Nuitka compiles all the constructs available in the reference implementation CPython – but promises a higher speed.
Commercial version for protection of IP
Nuitka is available in an open source version under an Apache 2.0 license on GitHub. Developers who have special requirements in the commercial environment, for example with regard to the protection of their intellectual property (Intellectual Property – IP), can alternatively use a paid Nuitka Edition. For fees starting at EUR 250 per year, this version offers additional plug-ins and targeted protection for the source code and any data integrated into the application.
If you want to delve deeper into the Python compiler, you can find more detailed insights in a live stream series by Nuitka developer Hayen on YouTube. In the recording of a lecture at PyConDE 2011, Hayen also explains his original motivation for developing the Python compiler.
Meanwhile, the blog post announcing the release summarizes a detailed overview of all changes and new functions in Nuitka 1.0.
