Skip to content

Latest commit

 

History

History
executable file
·
76 lines (56 loc) · 3.22 KB

README.md

File metadata and controls

executable file
·
76 lines (56 loc) · 3.22 KB

ros-docker-images

docker startup demo

ROS desktop-full docker images with Qt C++ debug tool-chain for Fedora and other Linux users who do not use Ubuntu.

This project aims to build an All-in-One development environment for robot learning including robotics and deep reinforcement learning and a portable platform for intelligent robot applications.

Features

  • integrated with open source AMD and Intel GPU driver, and NVIDIA driver manually installation instruction
  • support gazebo simulation
  • including QtCreator and RoboWare as ROS IDE which supports build and debug your ROS packages
  • including terminator, a multi-windows supported free terminal emulator
  • including ranger - a console file manager with VI key bindings
  • including ros-desktop-full packages, currently support indigo and kinetic version
  • including a light weight file manager GUI application pcmanfm

Usage

  • Pull docker image
$ docker pull jacknlliu/ros:kinetic-ide-init

See it on docker hub.

  • set X access on your host for docker containers
$ xhost +
  • Run
docker run --privileged  \
 --security-opt label=disable  \
 --security-opt seccomp=unconfined  \
 --env="DISPLAY" --env QT_X11_NO_MITSHM=1  \
 --volume="/tmp/.X11-unix:/tmp/.X11-unix:rw"  \
 --volume="/home/<your_user_name>/Workspace:/data:rw"  \
 --volume="/path/to/gazebo_models:/home/ros/.gazebo/models:rw"  \
 --name="ros_kinetic"  \
 jacknlliu/ros:kinetic-ide-init terminator

NOTE: /home/<your_user_name>/Workspace defines your workspace according to your own workspace to share between your host and docker container,

/path/to/gazebo_models defines your gazebo models directory path.

Run the above command just in the first time, later please just run the container with its name, this will keep things easy.

$ docker start ros_kinetic
  • Troubleshooting with NVIDIA video card
    If you use NVIDIA video card, you should follow this instruction.

  • Run ros docker images on multiple hosts, please refer to robot-manipulation-cluster.

Known Issues and Limitations

  • It's a little large image.
  • common issues about NVIDIA GPU driver

Reference

LICENSE

This project is distributed under MIT License.