A lightweight framework on top of the Qt scene graph and only few classes from Qt/Quick. It is usable from C++ and/or QML.
Go to file
Uwe Rathmann f49d99ac3c windows warning flags disabled - treats some warnings as errors 2023-04-05 18:47:43 +02:00
.github/workflows branch fixed 2023-04-05 18:37:55 +02:00
cmake windows warning flags disabled - treats some warnings as errors 2023-04-05 18:47:43 +02:00
doc focus handling removed ( spin boxes do not have an internal focus chain 2023-02-19 14:24:09 +01:00
examples bye,bye qmake, say hello to cmake 2023-04-05 17:47:15 +02:00
inputcontext bye,bye qmake, say hello to cmake 2023-04-05 17:47:15 +02:00
playground bye,bye qmake, say hello to cmake 2023-04-05 17:47:15 +02:00
qmlexport bye,bye qmake, say hello to cmake 2023-04-05 17:47:15 +02:00
skins bye,bye qmake, say hello to cmake 2023-04-05 17:47:15 +02:00
src bye,bye qmake, say hello to cmake 2023-04-05 17:47:15 +02:00
support bye,bye qmake, say hello to cmake 2023-04-05 17:47:15 +02:00
tools bye,bye qmake, say hello to cmake 2023-04-05 17:47:15 +02:00
.clang-format bye,bye qmake, say hello to cmake 2023-04-05 17:47:15 +02:00
.gitignore bye,bye qmake, say hello to cmake 2023-04-05 17:47:15 +02:00
.qmake.conf project file include mechanism based on QMAKEFEATURES 2019-03-31 13:05:25 +02:00
.uncrustify upgraded to uncrustify 0.76.0 - still not happy with the C++ beautifier 2023-02-28 15:49:42 +01:00
CMakeLists.txt bye,bye qmake, say hello to cmake 2023-04-05 17:47:15 +02:00
COPYING initial commit 2017-07-21 18:21:34 +02:00
README.md bye,bye qmake, say hello to cmake 2023-04-05 17:47:15 +02:00

README.md

QSkinny

CMake Build Matrix

The (Q)Skinny library is a framework built on top of the Qt scene graph and very few core classes from Qt/Quick. It offers a set of lightweight controls, that can be used from C++ and/or QML.

Doing the implementation in C++ allows to make use of the "scene graph" classes. Building controls from scene graph nodes allows for a lighter implementation than found with stacking "heavier" objects like QObject or QQuickItem. Offering a full featured C++ API allows the application code to benefit from following the same strategies.

Being "skinny" also means a design that separates concerns between the API and logic of the controls themselves, the styling of these controls, and the delegated rendering of the controls to the screen.

The code already provides a solid fundament for an automotive GUI with currently ~300K lines of pure C++ code. As expected it results in a good startup performance and a low memory footprint.

Nontheless QSkinny is lacking in areas like documentation or appealing default skins. The QML API has not been completed after reaching a proof of concept state. Furthermore the current selection of the implemented controls is limited to the needs of the driving projects.

QSkinny is supposed to run on all platforms being supported by Qt/Quick. But so far only Linux is actively tested. It might support all versions Qt >= 5.15, but you can rely on:

  • Qt 5.15
  • current long term supported ( LTS ) version of Qt
  • current version of Qt

On debian bullseye these packages need to be installed: build-essential qt-qmake qtbase5-dev qtbase5-private-dev qtdeclarative5-dev qtdeclarative5-private-dev libqt5svg5-dev.

Optional: When enabling the hunspell feature the following package needs to be installed: libhunspell-dev

Optional: When enabling the pinyin feature the following package needs to be installed: libimepinyin-dev

If you want to know more about QSkinny - or even like to give it a specific direction - please contact support@qskinny.org.

IOT dashboard

Contributing to QSkinny

For now QSkinny is licensed under LGPL v2.1. In the future, We want to have the possibility to transfer the QSkinny code to a more permissive license, i.e. any of the following or similar licenses: BSD "Simplified" License, BSD "Revised" License or MIT license.

If you are fine with these terms, you are welcome to contribute to QSkinny by completing the following steps:

  1. Agree to the terms of the Contributor Agreement. You can create a pull request first, then the licensing bot will check automatically whether you already signed the Agreement or not. If you have not signed it yet, it will take you to the Agreement to agree.
  2. Submit your pull request for review.