Nuitka and Google Summer of Nuitka

Google Summer of Code 2019

Nuitka team is happy to see your interest in helping out by applying for Google Summer of Code 2019 and helping develop it.

This is the first time Nuitka project participates, and we are excited to be joined by the students in this. We are experienced coders and have mentored during day jobs. And definitely we look forward to having fun with this.

About Nuitka

Nuitka is a project to write a Python compiler in Python. Unlike PyPy, it is about static compilation, and aims to be less complex, but fully compatible with gradual degradation of performance, up to C level.

See our introduction for a full summary and consider reading the user manual and the developer manual for further clarification.

What we do

The software created is a compiler that is a drop-in replacement for CPython and derivatives like Anaconda.

It aims at accelerating Python closer to C level while maintaining the full compatibility, while also delivering an option for deployment of applications in a standalone fashion.

Why is it interesting?

Python is used in a lot of places, but when it comes to performance, other languages are used, have to be used. Nuitka aims at changing it. Making Python even more of a choice for high performance computing will make us famous, rich, or loved. Maybe even all of that.

Who uses it?

Nuitka already has a lot of users. Many of these are most interested in the packaging side, but also some care about even tiny performance improvements, like the roughly 2x-3x speed up achieved in most cases.

What languages is it written in?

Nuitka is written in Python. It has a C run time, that for some ideas you would have to touch, for some you absolutely do not. There is a lot to be done in pure Python, esp. static optimization.

How is it going to change the world?

Being able to write C speed code in a language as simple as Python. What else to say. This is taking Python where it currently cannot be.

Contacting Nuitka

For the purpose of GSoC the Nuitka development mailing list can be used, as well as private email. We will be reachable via Hangout/Duo calls on a regular basis.

We will be able to accomodate all time zones somehow.

Getting Started

Nuitka setup is trivial and explained in the User Manual. Using Anaconda Python on Windows, or a normal Linux system, and you are good to go with a git clone of Nuitka in a minute. See Download page how to do that.

Bug fixes and features are expected to be done via PRs on Github. We can guide you through that, it's easy. We are not very formal, but helpful.

Remember, Nuitka must be the title of your application or else Google will not know what it is about.



1. Nuitka support for PyPI top 50

Project description: Nuitka works with most software. The aim of this project is to make sure it's true for the top 50 packages on PyPI, by compiling and using their example codes.

So, first of all, we need to define what we mean by top 50 PyPI packages. For Python downloading the packages via pip is very widespread. So the first task would be to look at the top 100 and identify 50 packages that are even relevant to Nuitka. There is a list of Top 100 PyPI.

Of course pip itself, leading the list, we will skip that, nobody needs that compiled or uses it as a library in their software. But urllib3 is currently top 2, and a very important piece of software that people rely on, so breaking it in Nuitka compilation would be bad.

What the student would be expected to do, is to go and read its documentation in search for a tutorial, small examples, that make it work. Distill a minimal test out of that and add it to the existing pool of tests in tests/standalone say as tests/standalone/ and get it to work in standalone mode, using the Nuitka argument --standalone.

Then, if possible and easy enough, the tests should be made to work. For hard cases, where the tests are difficult to get to work, we can skip them. For the widespread py.test and nose2 there is a very good chance to get them working against a compiled package. For that you would be using the Nuitka argument --package, throw in magic environment variables if necessary (this secret will be revealed, eh documented in the course of our work), if the tests live below the package name space, say urllib3.tests and then run the tests.

Looking at the test results, which hopefully pass and fail equally well (pro tip, never assume a released software passes all the tests when not compiled in your environment or any), you then try to identify the issue, or report it as a Nuitka issue, or sometimes as an issue of the software we are testing.

When those work, we should try and turn this into a re-usable test as well, so we can apply them in an automated fashion.

Then on to next package on the list. The main benefit to the student will be to get to know the 50 most important software packages of Python on at least a cursory level. Something the mentors won't even do. And that will teach you a lot and the mentors too. And it will prevent people using Nuitka from then encountering things that our testing will then find before our releases.

In a first stage, you would identify and report the issues to the bug tracker, in a second stage develop tools that help to narrow down issues, e.g. what extension module fails to load precisely, even with a segfault happening, and put them to use and try to fix a few of the simpler issues.

Setting up these as automated tests would be the ultimate goal, so we can follow these top 50 packages with Nuitka over time and make sure they continue to work. We can definitely help with that part though, the student may or may not have the ability to do that part himself, would be OK.

In the past it has happened e.g. that Jinja2 was breaking for Python 3.7, and it would be cool to discover this immediately.

Skills: Python programming, pip installation, virtualenv. Also need a Linux and/or Windows installs of Python, one platform is good, both would be great. Ideally learn about pipenv and apply it for defining environments to test in.

Main platform for this would be the Github issue and feel to ask questions and clarifications there.

Difficulty level: Easy

Potential mentors: Kay Hayen, Kamran Ghanaat, Jorj X. McKie

2. Nuitka one file standalone option

Project description: Nuitka has a mode meant for distribution to another system that puts everything needed in a single folder with a .dist suffix. This folder is then essentially the distribution.

One complaint often raised about that solution is that it's a folder rather than a single file, for alternative packaging methods, e.g. py2exe and pyinstaller, these do actually exist, and this project would be about integrating with that.

In a first stage, the student would identify the code of these tools that is doing it subsequently and try to port it to Nuitka for one or more platforms.

The main job here to analyse the competing projects code and to transfer the ability to Nuitka.

Skills: Python programming, having Linux and/or Windows installs of Python, both would be great. Likely some C knowledge may be required, but that is uncertain.

Main platform for this would be the Github issue and feel to ask questions and clarifications there.

Difficulty level: Easy

Potential mentors: Kay Hayen, Kamran Ghanaat

3. Nuitka benchmarks

Project description: Nuitka has too little in the way of measuring the actual performance gains one has. You would change that.

In a first stage, you would enhance the existing speedcenter to provide a more complete set of micro-benchmarks, for the different levels of optimization, with more or less type knowledge. You would then as a second step add a history of commits in some form of graphs that extend over a longer perioud of time, and automatically identify changes that e.g. produce equivalent C code.

As Nuitka is both about high level compile time optimization as well as low level runtime optimization, your task would be to enhance coverage and to make the information used to input decision making for Nuitka optimization more readable.

But also users should get a better grasp of what can be expected to be accelerated and what not, and by how much. The student will be relatively free in inventing ways to present this information.

Skills: Python programming, Linux installs of Python, C tooling would be nice, but can be mentored.

Main platform for this would be the Github issue and feel to ask questions and clarifications there.

Difficulty level: Intermediate

Potential mentors: Kay Hayen, Kamran Ghanaat, Jorj X. McKie

4. Nuitka all built-ins

Project description: Nuitka has support for many built-ins, e.g. len already, which means dedicated C code, compile time evaluation, type shapes produced (in this case an int), but there are some notable exceptions, e.g. enumerate where we know types too, that are still missing, but definitely can have high performance impact on some loops. Not having that means that enumerate using loops are loosing out on many optimization opportunities.

The students task would be to immitate existing built-in codes to achieve a complete support for ultimately all C built-ins. The first step would be to identify which ones are missing (by means of a warning added), then to find out in test runs of the test suites, which ones are warned about, and to resolve as many of those as possible. It is assumed that achieving this for all built-ins can be done with your help.

This would be great getting your feet wet with optimization in Nuitka and one that has actual impact, as well as seeing many corner cases of built-ins in Python that will the student will become knowledgable of.

Further reading:

  • Nuitka len node code

  • Runtime C code example:

    PyObject *BUILTIN_LEN(PyObject *value) {
        Py_ssize_t res = PyObject_Size(value);
        if (unlikely(res < 0 && ERROR_OCCURRED())) {
            return NULL;
        return PyInt_FromSsize_t(res);
  • Many more links and examples in the Github issue below.

Skills: Python and C programming, platform wouldn't matter

Main platform for this would be the Github issue and feel to ask questions and clarifications there.

Difficulty level: Intermediate

Potential mentors: Kay Hayen, Kamran Ghanaat

5. Nuitka macOS CI

Project description: Nuitka has currently no CI for macOS, which means it can be broken in any release.

Your task would be to enhance the Travis configuration to introduce that the tests are run on macOS too. Ideally you would also manage to get Anaconda on that platform used, but that is not expected.

Your mentors will not be able to help with macOS specifics. Nuitka is known to work on the platform, but Travis might expose differences that need some addressing.

Main platform for this would be the Github issue and feel to ask questions and clarifications there.

Skills: Travis, have macOS platform, XCode tooling

Difficulty level: Hard

Potential mentors: Kay Hayen, Kamran Ghanaat


This is time line as relevant for the students:

  • February 26 - organizations announced (PSF is going to be part of it)
  • Up to March 20 students discuss applications with mentoring organizations
  • March 25 - April 9th Student application period
  • May 6 Accepted student proposals announced
  • May 6 - May 27 community bonding
  • May 27 - Aug 26 coding
  • August September 3 results announced