Merge pull request #26 from Tiryoh/feature/docker-support

Support both nvidia/without nvidia

Dockerfile

@@ -21,7 +21,8 @@ WORKDIR /catkin_ws/src
 COPY . mycobot_ros
 
 # Install build dependencies
-RUN apt-get update && \
+RUN curl -s https://raw.githubusercontent.com/ros/rosdistro/master/ros.asc | apt-key add - && \
+    apt-get update && \
     apt-get install -y \
         # ROS Build dependencies
         python-rosinstall \

README.md

@@ -26,12 +26,20 @@ There are two ways to run this project. The first is by running the project in a
 [installing docker-compose](https://docs.docker.com/compose/install/). The benefit of running in the container is that you can run the project in any version of linux, as long as your kernel
 is new enough. 
 
-Once docker is installed, run the following command, and the project should show up:
+Once docker is installed, run the following command, and the project should show up.
+
+without NVIDIA GPU:
 
 ```
 docker-compose build ros && xhost +local:root && docker-compose up ros
 ```
 
+with NVIDIA GPU
+
+```
+docker-compose build nvidia-ros && xhost +local:root && docker-compose up nvidia-ros
+```
+
 This command does three things:
 1) `docker-compose build ros`
    
@@ -46,7 +54,7 @@ This command does three things:
 3) `docker-compose up ros`
 
    This runs the image specified in the `docker-compose.yml`, which by default runs
-   the command `roslaunch mycobot_ros control_slider.launch` within the container.
+   the command `roslaunch mycobot_ros mycobot_slider.launch` within the container.
    
 
 ### Option 2: Local

docker-compose.yml

@@ -6,7 +6,7 @@ x-app: &common
   environment:
     PYTHONUNBUFFERED: 1
     QT_X11_NO_MITSHM: 1
-    DISPLAY: :1
+    DISPLAY: $DISPLAY
   volumes:
     - "/tmp/.X11-unix:/tmp/.X11-unix:rw"
   build:
@@ -19,7 +19,9 @@ x-app: &common
 
 services:
   # This configuration builds the project for computers with NVIDIA hardware
-  nvidia-ros: *common
+  nvidia-ros: 
+    <<: *common
+    runtime: nvidia
 
   # This is the default for computers that don't have nvidia hardware accelerators
   ros: