Skip to content


Repository files navigation

QtQuick Virtual Control Panel

A remote UI implementation for Machinekit written in Qt/C++/QML. More info...

Travis CI AppVeyor
Build Status Build Status Build Status

Coverity Scan Build Status


  • enables building remote user interfaces for Machinekit
  • automatic service discovery
  • control one or multiple Machinekit instances
  • remote UI deployment via MachinekitClient
  • remote Machinekit instance launcher
  • UI components for machine control applications
  • textual GCode preview
  • 3D GCode preview
  • QML live-coding support
  • cross-platform: runs on Windows, Linux, Mac OS X, Android, and iOS


Binaries for the QtQuickVcp library and MachinekitClient can be found here:

From App Store

The MachinekitClient is available in following app stores:

Store Link
Google Play Store Get it on Google Play!

Rolling releases

You can find rolling releases of the MachinekitClient and the QtQuickVcp library below. Stable releases have been dropped in favor of continuous deployment to decrease maintenance effort.

Operating System MachinekitClient QtQuickVcp
Windows 64bit Download Download
Windows 32bit Download Download
Linux 64bit Download Download
Mac OS X 64bit Download Download
Android armv7 Download Download

QtQuickVcp direct download links

For integration into your CI system, QtQuickVcp can be directly downloaded using a simple shell script. For example, the latest QtQuickVcp Android build can be downloaded with the following command:

package=$(wget -qO- | grep ${arch} | grep ${extension} | grep ${platform} | tail -n 1 | awk -F"\"" '{print $4}')
wget -O qtquickvcp.${extension} ${url}


What is QtQuickVcp?

What is the MachinekitClient?

Showcase Applications and Examples

Getting Started

Building and Installing


Developing QtQuickVcp

What is QtQuickVcp?

QtQuickVcp is a Machinekit component which adds the ability create new user interfaces for Machinekit. Unlike GladeVCP all user interface functionality can be implemented in QML only.

QtQuickVcp is designed with the aspect of remote interfaces on mobile devices and embedded systems in mind. A UI is designed once for a specific form factor and can then be deployed to any mobile, embedded or desktop device over the network.

What is the MachinekitClient?

The MachinekitClient application is intended to be used as a generic client for QtQuickVcp based user interfaces. MachinekitClient serves as client for remote distribution of QtQuickVcp-based user interfaces.

Using the generic MachinekitClient is the easiest way to use QtQuickVcp. However, you can also build QtQuickVcp UIs as standalone applications.

Showcase Applications and Examples

Full-featured UIs

Please add your awesome QtQuickVcp application here!

Example UIs

Machinekit configurations

Getting Started

A good way to get started is to watch the video tutorials on YouTube.

Another resource for learning about QtQuickVcp, Machinekit, and other machine control related stuff is Machine Koder Blog

HelloWorld in QtQuickVcp

HelloWorld in QtQuickVcp

Building Qt5 UIs with MachinekitSDK

Building Qt5 UIs with MachinekitSDK

More detailed videos for specific areas will follow.

QtQuick Module Overview

QtQuickVcp comes with the following QtQuick modules.

  • Machinekit.HalRemote - Non GUI HAL remote pins and components
  • Machinekit.HalRemote.Controls - GUI items such as Slider and ProgressBar combined with HAL pins
  • Machinekit.Application - Non GUI CNC stack application components
  • Machinekit.Application.Controls - GUI items for CNC type applications
  • Machinekit.Controls - Generic UI items not combined with HAL pins (can be used for non HAL projects)
  • Machinekit.PathView - GCode path and progress views
  • Machinekit.VideoView - Video views, e.g. for mjpeg-webcam streams

Using mkwrapper

Mkwrapper is the Machinekit server side application for the Machinekit.Application modules. This short intro will help you making your Machinekit configuration work with mkwrapper.

Install or Update Machinekit

If you have no Machinekit installation please follow the installation steps in the wiki Debian Packages

If you already have Machinekit installed make sure it is up to date:

sudo apt-get update
sudo apt-get upgrade

In some cases, you need a dist-upgrade to upgrade packages from non-debian repositories.

sudo apt-get dist-upgrade

NOTE: When executing the dist-upgrade, make sure you have only the Debian and the Machinekit repositories activated. Alternatively, you can update the relevant Machinekit packages manually by running sudo apt-get install <machinekit packages>.

Prepare Machinekit

To enable remote communication you have to set REMOTE variable to 1 in the ini-file:

sudo nano /etc/linuxcnc/machinekit.ini

NOTE: Enable remote communications only in a secured private network. At the moment Machinetalk has no security layer.

Prepare your Configs

To use mkwrapper and Machinetalk for your existing configurations you have to do 3 modifications.

Modify the INI-file

Linuxcnc needs to know which user-inferface it should use. For mkwrapper you need to edit following in the DISPLAY section:

DISPLAY = mkwrapper
Modify the HAL-file

For some user-interface you need a running Haltalk server. You can add one to you existing configuration by adding following in the end of the HAL file:

# start haltalk server
loadusr -W haltalk

Note: It is important to add the above line to the end of the HAL file, not in the beginning or else you will encounter connection problems.

Create a Run-Script

Machinekit configurations need a few actions to be performed before linuxcnc can start. Take a look at the following script and modify it to fit your configuration.

Name it <name of your configuration>.py and run chmod +x <filename> to make it executable.

NOTE: The example run script contains commands that you may not need for your specific setup. Please read the comments carefully and uncomment and modify commands that you may need for your setup.


import sys
import os
import subprocess
import importlib
from machinekit import launcher
from time import *


    launcher.check_installation()  # make sure the Machinekit installation is sane
    launcher.cleanup_session()  # cleanup a previous session
    # Uncomment and modify the following line if you create a configuration for the BeagleBone Black
    # launcher.load_bbio_file('myoverlay.bbio')  # load a BeagleBone Black universal overlay file
    # Uncomment and modify the following line of you have custom HAL components
    # launcher.install_comp('gantry.comp')  # install a comp HAL component if not already installed
    launcher.start_process("configserver ~/Machineface ~/Cetus/")  # start the configserver with Machineface an Cetus user interfaces
    launcher.start_process('linuxcnc xyz.ini')  # start linuxcnc
except subprocess.CalledProcessError:

# loop until script receives exit signal
# or one of the started applications exited incorrectly
# cleanup is done automatically
while True:
Modifications for 3D Printers

The Machineface user interface is especially designed for controlling 3D printers. Therefore it features some additional controls for heatbed, extruder temperature and more. It implements following a set of MCodes as well as a few HAL signals. You may also want to take a look at the source code of Machineface: DisplayPanel.qml

Cloning the User Interfaces

At the moment two user interfaces based on QtQuickVcp exist:

Machineface: a user interfaces designed for the use with small screens especially for 3D printers

Cetus: a full-featured user interfaces that should provide something like Axis

Clone both user interfaces on your device with Machinekit installed. I recommend forking the user interfaces on GitHub and cloning the fork so you can push modifications upstream quickly. To modify these user interfaces a simple text editor is sufficient. Using a text editor with syntax highlighting for QML might improve the development experience. (e.g. Kate or Qt Creator)

git clone
git clone

Install the MachinekitClient

Please download and install the MachinekitClient from the Download section.

Test it

Now its time to start your configuration:

./<name of your start script>.py

It may be useful to start the configuration at boot: Starting a Machinekit configuration at boot

Now start the MachinekitClient on the desired platform. For some networks it may be necessary to enter the IP address of the Machinekit device manually (Unicast).

Setup mklauncher

With newer versions of QtQuickVcp the entry point for the MachinekitClient has been changed to the mklauncher service.

Using mklauncher

Mklauncher is the new entry point for MachinekitClient. It acts as a remote version of the Machinekit Launcher. Usage is quite simple (use =mklauncher --help= for details) the only thing you need to add to your Machinekit configarion is a =launcher.ini= file.

name = MendelMax CRAMPS
description = DIY 3D printer
command = python
variant = with one extruder

More examples can be found here MendelMax-CRAMPS and here thecooltool-config.

Once you have create a launcher.ini file and you configuration is ready for testing you can start mklauncher by typing following command:

mklauncher .

The dot in the end means that mklauncher will recursively search for launcher.ini files in the current directory. NOTE that this can be very slow if you do this in your home directory.

Once you have successfully launched mklauncher you are ready to connect using the MachinekitClient.

If you have a embedded Machinekit setup e.g. on the BeagleBone Black, it is recommended to start mklauncher at boot using systemd. Use this guide for reference: Starting a Machinekit configuration at boot or use this script

Using videoserver

QtQuickVcp supports webcam streaming. Webcam streaming uses the videoserver as entry point. The videoserver requires mjpeg-streamer in a special version with ZeroMQ support to be installed on the system.


To install mjpeg-streamer on Debian Stretch use the following commands on you Linux computer:

sudo apt update
sudo apt install libzmq3-dev protobuf-c-compiler

Then clone and build mjpeg-streamer

git clone
cd mjpeg-streamer
make -C mjpg-streamer-experimental
sudo make -C mjpg-streamer-experimental install


Videoserver requires a video.ini file for creating new video streaming instances:

framerate: 30
resolution: "320x240"
quality: 80
device: "/dev/video0"
bufferSize: 1

To start videoserver run:

videoserver -i video.ini Webcam1

To automatically start videoserver from you you can use following snippet:

if os.path.exists('/dev/video0'):  # automatically start videoserver
    launcher.start_process('videoserver -i video.ini Webcam1')

Building and installing

QtQuickVcp is very versatile and is available for following platforms:

  • Windows Vista or newer
  • OS X Mavericks or newer
  • Linux (tested on Debian Jessie)
  • Android 4.x or newer
  • iOS (no binaries yet)

Please note that intalling QtQuickVcp can be very cumbersome for most platforms. However, since QtQuickVcp use remote deployment of it is not necessary that you build QtQuickVcp e.g. for Android. Instead please take the easy way and use the Machinekit Vagrant box for development. MachinekitClient serves as universal client application. You can download binaries for all supported platforms. Please only try to build and install QtQuickVcp on you own if you feel confident to do so and if you plan to contribute to the QtQuickVcp project.

Easy Way - Vagrant Box

The easiest way to get a running MachinekitSDK, QtQuickVcp and Machinekit installation is to use the Vagrant configuration. Follow the steps provided in the repository and you will have a working installation with a few clicks.

Generic Requirements

QtQuickVcp has the following requirements:

  • Qt SDK with Qt 5.4.0 or newer Note that Qt 5.4.2 to Qt 5.5.1 will not work on Linux
  • Protocol Buffers - version 2.5.1 or newer
  • ZeroMQ - version 3.x or newer
error C2338: <hash_*> is deprecated and will be REMOVED. Please use <unordered_*>.
You can define _SILENCE_STDEXT_HASH_DEPRECATION_WARNINGS to acknowledge that you have received this warning.

If MSVC2015 is the only option you have, add _SILENCE_STDEXT_HASH_DEPRECATION_WARNINGS=1; to "Preprocessor Definitions" entry under Project Properties. See also C++ Hash Deprecation Warning


Build instruction for Android toolchain on Linux


  • Install Qt SDK for Android
  • Download and extract Android NDK and Android SDK to ~/bin
  • Install dependency packages: sudo apt install -y libtool-bin make curl file libgtest-dev python default-jdk ant lib32z1 lib32ncurses5 lib32stdc++6 python-pip

Stand-alone Android toolchain

First create a Android Stand-alone toolchain:

sudo ~/bin/android-ndk/build/tools/ --install-dir=/opt/android-toolchain --arch=arm
export PATH=/opt/android-toolchain/bin:$PATH


Not yet necessary. You can skip this step.

export OUTPUT_DIR=/opt/libsodium-android

git clone
cd libsodium
git checkout 1.0.9
./configure --enable-static --disable-shared --prefix=$OUTPUT_DIR
sudo make install


Alter and execute the following commands

export OUTPUT_DIR=/opt/zeromq-android
export RANLIB=/opt/android-toolchain/bin/arm-linux-androideabi-ranlib

git clone
cd zeromq4-x/
git checkout v4.0.8

# fix compile problems
mv tools/curve_keygen.c tools/curve_keygen.cpp
sed -i 's/\.c\>/&pp/' tools/
rm -f tools/.deps/curve_keygen.Po

./configure --enable-static --disable-shared --host=arm-linux-androideabi --prefix=$OUTPUT_DIR \
sudo make install


export PATH=/opt/android-toolchain/bin:$PATH
export CFLAGS="-fPIC -DANDROID -nostdlib"
export CC=arm-linux-androideabi-gcc
export CXX=arm-linux-androideabi-g++
export NDK=~/bin/android-ndk
export SYSROOT=$NDK/platform/android-9/arch-arm
export OUTPUT_DIR=/opt/protobuf-android

# Latest and greatest, you might prefer v2.5.0 for Debian wheezy and v3.0.2 for Debian stretch
# sincee it is usually installed in your distro
git clone
cd protobuf
git checkout v3.0.2

mkdir -p gtest/msvc
touch gtest/msvc/foo.vcproj
./configure --enable-static --disable-shared --host=arm-eabi --with-sysroot=$SYSROOT CC=$CC CXX=$CXX --enable-cross-compile --with-protoc=protoc LIBS="-lc" --prefix=$OUTPUT_DIR
sudo make install

Mac - OS X and iOS


Then run

sudo port selfupdate
sudo port install libtool automake m4 autoconf pkgconfig



Install ZeroMQ to /opt/local

git clone
cd zeromq4-x
git checkout v4.0.7
./configure --disable-static --enable-shared --prefix=/opt/local CC=clang CXX=clang++ CXXFLAGS="-std=c++11 -stdlib=libstdc++ -O3" LDFLAGS="-stdlib=libstdc++"
sudo make install

Installs ZeroMQ libraries for iOS to /opt/zeromq-ios

git clone
cd libzmq-ios
chmod +x
sudo ./



Since Yosemite one needs to compile a protobuf library that is compatible with libstdc++.

git clone
cd protobuf
git checkout v2.6.1
./configure --disable-shared --enable-static --prefix=/opt/local CC=clang CXX=clang++ CXXFLAGS="-std=c++11 -stdlib=libstdc++ -O3" LDFLAGS="-stdlib=libstdc++"
sudo make install

See installs protobuf to /opt/protobuf-ios

curl --output
chmod +x
sudo ./


The following steps are tested on Debian Jessie. For other distributions please use the äquivalent packages if available.

VirtualBox users see Machinekit Vagrant. Do not enable 3D acceleration or OpenGL will not work inside the VM.


sudo apt-get update
sudo apt-get install build-essential gdb dh-autoreconf libgl1-mesa-dev libxslt1.1 git
Protobuf and ZeroMQ packages

Debian Stretch

sudo apt install libprotobuf-dev protobuf-compiler libzmq5-dev

Debian Jessie

sudo apt-key adv --keyserver --recv 43DDF224
sudo sh -c \
"echo 'deb jessie main' > \
sudo apt-get update
sudo apt-get install pkg-config libprotobuf-dev protobuf-compiler libzmq3-dev
Protobuf from source
git clone
cd protobuf
git checkout v2.6.1
sudo make install
ZeroMQ from source
git clone
cd zeromq4-x/
git checkout v4.0.7
./configure     # add other options here
make check
sudo make install


If you want to use Windows in a VirtualBox VM please enable 3D acceleration for Qt to work properly.


  • Install everything to C:\bin
  • Install Microsoft Visual Studio
  • Community Edition
  • Install Qt SDK for Windows (Use the MSVC2015 or MSVC2013 version depending on Visual Studio)
  • Install a Git command line client (VS2015 comes with Git, Git for Windows is also fine, select add to Windows Path in setup)
  • Create a link from your users directory to C:\bin
  • Run in cmd window as administrator: mklink /D C:\Users\%USERNAME%\bin C:\bin


Open a cmd window:

cd C:\bin
git clone
cd zeromq4-x
git checkout v4.0.8

Now start Visual Studio and open the solution C:\zeromq4-x\builds\msvc\msvc11.sln say yes to migrating the project to the new format.

Wait a few seconds until parsing the header files is finished. Then select the Release build.

Now right click on the libzmq project in the Solutions Explorer and click build.

Now select the Win32 build.

Now right click on the libzmq11 project in the Solutions Explorer and click build.

Copy the libzmq.dll and libzmq.pdb files from zeromq4-x\bin\Win32 to the Qt\<version>\<release>\bin folder.

Same for x64


Open a cmd window:

cd C:\bin
git clone
cd protobuf
git checkout v2.6.1

Now start Visual Studio and open the solution C:\protobuf\vsprojects\protobuf.sln say yes to migrating the project to the new format.

Wait a few seconds until parsing the header files is finished. Then select the Release build.

Now right click on the libprotobuf project in the Solutions Explorer and click build. Also see MSVC2015

Repeat this step for libprotobuf-lite, libprotoc and protoc.

Copy the libprotobuf.lib files from protobuf\vsprojects\Release to the Qt\<version>\<release>\lib folder.

Copy the libprotobuf.pdb files from protobuf\vsprojects\Release to the Qt\<version>\<release>\bin folder.

Repeat for x64 in protobuf\vsprojects\x64\Release.

MSVC2015 and MSVC2017

Even if you have the option to use MSVC2015, use MSVC2013 instead by changing the platform tool set to Visual Studio 2013, as you will get this kind of compiler error with MSVC2015:

error C2338: <hash_*> is deprecated and will be REMOVED. Please use <unordered_*>.
You can define _SILENCE_STDEXT_HASH_DEPRECATION_WARNINGS to acknowledge that you have received this warning.

If MSVC2015 is the only option you have, add _SILENCE_STDEXT_HASH_DEPRECATION_WARNINGS=1; to "Preprocessor Definitions" entry under Project Properties. See also C++ Hash Deprecation Warning

### Building QtQuickVcp

After you have all the requirements installed clone and build the QtQuickVcp repo.

git clone

Before building the project modify the paths.pri file if necessary.

Build from Qt Creator

Open Qt Creator and open the file. Select the Qt version you want to build against. Before building the project add these additional make commands to your project settings: install, docs and install_docs Qt Creator Build Settings

Note: for android you need to prefix the install command as follows: -C 3rdparty install and -C src install else you will get errors during the Android "make install".

Build the project in release mode (or debug mode if you want to debug QtQuickVcp and you application). Now you should have a working QtQuickVcp installed to your Qt version.

Setup Qt Creator

Per default Qt Creator's QtQuick designer does not work with custom QML modules. Therefore, it is necessary to enable building a working QML emulation layer. This can be done in the Qt Creator preferences Tools > Options... in the Qt Quick register by selecting the Use QML emulation layer that is built by the selected Qt option. The build path is automatically completed.

Qt Creator QtQuick designer setup

Build from command line (Linux only)

The following steps assume that you have the Qt SDK installed in ~/bin/Qt/

# set QMAKE

# download and install QtQuickVcp
git clone
mkdir -p build/QtQuickVcp
cd build/QtQuickVcp

$QMAKE ../../QtQuickVcp
# to build without plugins.qmltypes use:
make docs
make install
make install_docs


If you also want to have QtQuickVcp specific wizards and extensions for Qt Creator please continue with the install instructions for the MachinekitSDK


  • Create an issue in the issue tracker (e.g. Fan control missing)
  • Fork the git repository.
  • After you have coded some cool new stuff please create a pull request and link it to the issue.

Developing QtQuickVcp

This wiki page is for developers who want to extend the functionality of QtQuickVcp.

If you are new to Qt and/or QtQuick I would recommend you to read the Qt documentation and follow the getting started and tutorials:

Coding Style Guidelines

If you want to contribute to the project please respect the coding style guidelines of the project outlined in the .editorconfig file.


Unit testing in this project is powered by Catch and Trompeloeil.

Please take a look at the provided links for more information on how to use the unit testing and mocking frameworks.

Choosing the right QML module

QtQuickVcp is split into several QML modules:

  • Machinekit.HalRemote - Non GUI HAL remote pins and components
  • Machinekit.HalRemote.Controls - UI items such as Slider and ProgressBar combined with HAL pins
  • Machinekit.Controls - Generic UI items not combined with HAL pins (can be used for non HAL projects)
  • Machinekit.PathView - GCode path and progress views
  • Machinekit.VideoView - Video views, e.g. for mjpeg-webcam streams

You can find these modules as separate folders inside the src folder of QtQuickVcp. E.g. the folder containing the module Machinekit.HalRemote is src/halremote

Creating new QML based Components

QML files are the preferred way of creating new visual QtQuickVcp Components. Just create a new QML file using the New File dialog and then you can either use the Qt Quick Desinger or the text editor to develop your components. I recommend you to use the text editor as you will learn how to use make use of QtQuick faster and the Designer is currently far from perfect.

A good way to start is looking at the existing QML Components and how they are implemented. Please make sure you make use of[anchors or layout] based positioning wherever possible as it saves computation power and makes your UIs scaleable.

Adding QML files to the QML module

You need to add the newly created QML files to following places in order to make them work when the QML module is deployed.

  • .pro file of the module to the QML_FILES variable
  • .qrc file of the module
  • plugin.cpp of the module to the qmldir structure
  • .metainfo file in designer folder

Creating new C++ based Components

For some functionality it is necessary to develop C++ based QtQuickVcp Components. This may include visual items that need access to native OpenGL painting instructions, visual items that use QPainter to paint on a canvas and non visual items that need performance, access to Qt/C++ functionality or native C/C++ libraries.

Adding C++ classes to the QML module

You need to add the C++ classes to following places in order to make them work when the QML module is deployed.

  • plugin.cpp of the module with qmlRegisterType
  • .metainfo file of the Qt Quick Designer plugin

Creating a new QML module


Adding the QML module to the project

  • (OTHER_FILES, doc folder)
  • src/
  • doc/config/qtquickvcp-project.qdocconf (sourcedirs)