Table Of Contents

• Installation for Developement
  • Development Version
    • Installing Dependencies
      • Linux
      • Ubuntu
      • OS X
      • OSX HomeBrew
      • Windows
    • Installing Kivy for Development
    • Running the test suite

Installation for Developement

Development Version

Note that when running a development version, you’re running potentially broken code at your own risk. To use the development version, you will first need to install the dependencies. Thereafter, you will need to set up Kivy on your computer in a way that allows for easy development. For that, please see our Contributing document.

Installing Dependencies

Linux

To install Kivy’s dependencies, follow the guide below for your platform. You will also need these packages for RST, lexing and image handling:

$ python -m pip install pygments docutils pillow --user

Ubuntu

For Ubuntu 12.04 and above (tested to 14.04), simply enter the following command that will install all necessary packages:

$ sudo apt-get install python-setuptools python-pygame python-opengl \
  python-gst0.10 python-enchant gstreamer0.10-plugins-good python-dev \
  build-essential libgl1-mesa-dev-lts-quantal libgles2-mesa-dev-lts-quantal\
  python-pip

For Ubuntu 15.04 and versions older than 12.04, this one should work:

$ sudo apt-get install python-setuptools python-pygame python-opengl \
  python-gst0.10 python-enchant gstreamer0.10-plugins-good python-dev \
  build-essential libgl1-mesa-dev libgles2-mesa-dev zlib1g-dev python-pip

For Ubuntu 16.04:

$ sudo apt-get install python-setuptools python-pygame python-opengl \
  python-gst0.10 python-enchant gstreamer0.10-plugins-good python-dev \
  build-essential python-pip libgl1-mesa-dev libgles2-mesa-dev zlib1g-dev

For Ubuntu 17.10 and 18.04:

$ sudo apt-get install python-setuptools python-pygame python-opengl \
  python-enchant python-dev build-essential python-pip libgl1-mesa-dev \
  libgles2-mesa-dev zlib1g-dev

Kivy requires a recent version of Cython, so you might need to update your Cython installation.

$ python -m pip install --user --upgrade Cython==0.29.9

If you experience compilation problems, please check your version of Cython against those in the Cython dependencies section.

OS X

Without using brew you can install the dependencies for kivy by manually pasting the following commands in a terminal:

curl -O -L https://www.libsdl.org/release/SDL2-2.0.9.dmg
curl -O -L https://www.libsdl.org/projects/SDL_image/release/SDL2_image-2.0.1.dmg
curl -O -L https://www.libsdl.org/projects/SDL_mixer/release/SDL2_mixer-2.0.1.dmg
curl -O -L https://www.libsdl.org/projects/SDL_ttf/release/SDL2_ttf-2.0.13.dmg
curl -O -L http://gstreamer.freedesktop.org/data/pkg/osx/1.7.1/gstreamer-1.0-1.7.1-x86_64.pkg
curl -O -L http://gstreamer.freedesktop.org/data/pkg/osx/1.7.1/gstreamer-1.0-devel-1.7.1-x86_64.pkg
hdiutil attach SDL2-2.0.9.dmg
sudo cp -a /Volumes/SDL2/SDL2.framework /Library/Frameworks/

This should ask you for your root password, provide it and then paste the following lines in your terminal:

hdiutil attach SDL2_image-2.0.1.dmg
sudo cp -a /Volumes/SDL2_image/SDL2_image.framework /Library/Frameworks/
hdiutil attach SDL2_ttf-2.0.13.dmg
sudo cp -a /Volumes/SDL2_ttf/SDL2_ttf.framework /Library/Frameworks/
hdiutil attach SDL2_mixer-2.0.1.dmg
sudo cp -a /Volumes/SDL2_mixer/SDL2_mixer.framework /Library/Frameworks/
sudo installer -package gstreamer-1.0-1.7.1-x86_64.pkg -target /
sudo installer -package gstreamer-1.0-devel-1.7.1-x86_64.pkg -target /
python -m pip install --upgrade --user Cython==0.29.9 pillow

Now that you have all the dependencies for kivy, you need to make sure you have the command line tools installed:

xcode-select --install

Go to an appropriate dir like:

mkdir ~/code
cd ~/code

You can now install kivy itself:

git clone http://github.com/kivy/kivy
cd kivy
make

This should compile kivy, to make it accessible in your python env just point your PYTHONPATH to this dir:

export PYTHONPATH=~/code/kivy:$PYTHONPATH

To check if kivy is installed, type the following command in your terminal:

python -c "import kivy"

It should give you an output similar to the following:

$ python -c "import kivy"
[INFO   ] [Logger      ] Record log in /Users/quanon/.kivy/logs/kivy_15-12-31_21.txt
[INFO   ] [Screen      ] Apply screen settings for Motorola Droid 2
[INFO   ] [Screen      ] size=480x854 dpi=240 density=1.5 orientation=portrait
[INFO   ] [Kivy        ] v1.9.1-stable
[INFO   ] [Python      ] v2.7.10 (default, Oct 23 2015, 18:05:06)
[GCC 4.2.1 Compatible Apple LLVM 7.0.0 (clang-700.0.59.5)]

OSX HomeBrew

If you prefer to use homebrew: install the requirements using homebrew:

$ brew install sdl2 sdl2_image sdl2_ttf sdl2_mixer gstreamer

Windows

See Use development Kivy.

Installing Kivy for Development

Now that you’ve installed all the required dependencies, it’s time to download and compile a development version of Kivy:

Download Kivy from GitHub:

$ git clone git://github.com/kivy/kivy.git
$ cd kivy

Compile:

$ python setup.py build_ext --inplace -f

If you have the make command available, you can also use the following shortcut to compile (does the same as the last command):

$ make

Warning

By default, versions 2.7 to 2.7.2 of Python use the gcc compiler which ships with earlier versions of XCode. As of version 4.2, only the clang compiler is shipped with XCode by default. This means that if you build using XCode 4.2 or above, you need to ensure you have at least Python 2.7.3 installed, but preferably the latest version (2.7.5 at the time of writing).

If you want to modify the Kivy code itself, set up the PYTHONPATH environment variable to point at your clone. This way you don’t have to install (setup.py install) after every tiny modification. Python will instead import Kivy from your clone.

This simplest way to do this, is to install it as an editable install with:

python -m pip install -e .

And then the PYTHONPATH does not need to be set.

Alternatively, if you don’t want to make any changes to Kivy itself, you can also run (as admin, e.g. with sudo):

$ python setup.py install

If you want to contribute code (patches, new features) to the Kivy codebase, please read Contributing.

Running the test suite

To help detect issues and behaviour changes in Kivy, a set of unittests are provided. A good thing to do is to run them just after your Kivy installation, and every time you intend to push a change. If you think something was broken in Kivy, perhaps a test will show this. (If not, it might be a good time to write one.)

Kivy tests are based on pytest, which you can install from your package manager or using pip:

$ python -m pip install pytest

To run the test suite, do:

$ make test