Setting Up Your Development Environment

Table of Contents

1. Things You'll Need
2. Quick Docker Command Summary
3. Getting the ROS Docker VNC Image (technically optional)
   1. Pulling the Image
4. Running the Container
5. Entering the Container
   1. Desktop Interface
   2. Terminal Interface
   3. Editing with VSCode

P.S.: Please use NeoVim, it's awesome and I (Arvin) built on an excellent configuration here. Also, if you use emacs, you're dead to me /j.

A lot of this guide is completed in a Docker container. I know some nerd's gonna say something along the lines of "why would you ever use a Docker container as a development environment," and to them I say, "leave me alone, I'm new to Docker."

Now, if you want to guide to how to use Docker, check out the next subchapter, (after you get Docker set up, of course). It should be noted, however, that the rest of the guide will directly use Docker commands without too much explanation about why exactly we use those commands. Also, if this guide is not sufficient, feel free to google, I'm by no means an expert in Dockering.

Feel free to just skip this entire page if you already has ROS Noetic.

Things You'll Need

Sorry, I didn't warn you this page was structured like a cookbook /j.

• Pineapples
• A Docker installation (see here)
  • Note: If you're using Windows, I highly recommend the WSL version of installation.
• A decent internet connection (the ability to download about 4GB of data.)

Before proceeding, it may be useful to check out the next subchapter.

Quick Docker Command Summary

Note, command options marked in square brackets ([]) are optional. Additionally, command options marked with angle brackets (<>) customizable. For example, command [-a] <name> means that both command -a hypotenuse and command pineapple are valid commands.

Getting the ROS Docker VNC Image (technically optional)

If you head to our GitHub project embedml/docker-ros-desktop-vnc, you'll find more thorough instructions.

Pulling the Image

Let's start by grabbing our Docker image from hub.docker.com. This can be done with the following command:

$ docker pull arvinskushwaha/ros-noetic-desktop-vnc

If you get some sort of error that looks like this:

Got permission denied while trying to connect to the Docker daemon socket at unix:///var/run/docker.sock: Get "http://%2Fvar%2Frun%2Fdocker.sock/v1.24/containers/json?all=1": dial unix /var/run/docker.sock: connect: permission denied

Don't worry, it's just a permissions error. You'll have to rerun the command with sudo as follows:

$ sudo docker pull arvinskushwaha/ros-noetic-desktop-vnc

Same usually goes for any future errors of that form. More information on this can be found here. If you must use VSCode to edit within the container, you may have to do some wizardy from the aforementioned link to get the sudo prefix to not be required.

Now, when you do this command, you'll see something that looks like the following:

Running the Container

From our Docker image, we can build a container. This is done with the following command (do not run the following command yet):

$ docker run --name <container-name> -d arvinskushwaha/ros-noetic-desktop-vnc

The above command is essentially the base command which we add modifiers to. Also, where it says <container-name>, replace it with a name of your choice.

For these options, it is important to remember that they cannot be changed without just deleting and recreating the container. For tons of information on the options, check here.

Here are what the current command-line options mean:

• --name <container-name>: Each container is given a unique ID, and this option lets you customize it. It is important to note that this value must not conflict with other containers. This will be the identifier by which you reference a container.
• -d: This tells Docker to launch the container "detached". Without this option your terminal would simply freeze, as the container's main process just sleeps indefinitely.

Other command-line options you should know of:

• --net=<network>: This tells Docker to use a different network than the default (bridge). Values are typically host, bridge, container, or none. You can also create custom network configurations with Docker, but I am definitely not knowledgeable enough to say anything about that. Essentially, if you set --net=host, you never have to worry about container and host ports.

• -p <host-port>:<container-port>: Now, to worry about container and host ports. The VNC/NoVNC servers the container runs use port 5900 and 80 respectively. This means we can control how they are exposed on the host-side with the -p option, or just use them as-is with the --net=host option. Additionally, if there are other servers running inside the container, additional -p options may be added for those. As an example, if I wanted to map the VNC server port to port 3000 on localhost, I would add -p 3000:5900 as an option.

• --device=<host-path>:<container-path>[:modifier]: This option allows you to pick and chose which devices are passed through to the container. Since each device is a file in linux, we must pass in the device paths. For example, /dev/sda. If I wanted to map joystick 0 (/dev/js0) from the host to the container, I would add --device=/dev/js0:/dev/js0 as an option. I don't know must about the modifiers, so I'll just link the reference here.

• --privileged: This option is kind of dangerous. It, for all intents and purposes, gives full access to your device from the container. This can be useful to passthrough any device needed. See the reference above.

For my default setup, I'm going with the following:

$ docker --name rosenv -p 80:80 -d arvinskushwaha/ros-noetic-desktop-vnc

This allows me to just access the NoVNC server by typing in localhost in my browser.

Entering the Container

Now that we've made our container, it would be useful to be able to actually do things in it. There are two main ways of interacting with the container: terminal and desktop.

Desktop Interface

If you would like to use the desktop interface, you must have mapped either the VNC or NoVNC server ports such that it's accessible to the local host.

Then, for VNC, you can download some form of VNC server and connect to localhost:<mapped-vnc-port>. Or, in the case of NoVNC, just open localhost:<mapped-novnc-port> in the browser.

This is what it looks like:

Terminal Interface

If you would prefer using the terminal, execute the following command:

$ docker exec -it <container-name> bash

This will open a shell prompt that looks something like the following:

root@67c3c27e7401:/root#

Now, you may enter ./login, a nice simple login script we've written, or just do cd /home/ubuntu; su ubuntu -.

root@67c3c27e7401:/root# ./login

This shows yet another prompt:

root@67c3c27e7401:/root# ./login
ubuntu@67c3c27e7401:~$

To exit, execute exit twice. The first time should exit back to root and the second should exit the container.

root@67c3c27e7401:/root# ./login
ubuntu@67c3c27e7401:~$ exit
exit
root@67c3c27e7401:/root# exit
exit

Great! Now that we've got that down, let's talk about our in-container development environment.

In the home directory (/home/ubuntu), you will find two files and a directory:

• install_dev_tools: This file is run when the image was originally built. It contains information about the development tools installed on the container, i.e. NeoVim. This makes it easy for those who choose to primarily use the terminal to edit their files.

• upgrade: This script essentially just updates Ubuntu packages. It's recommended to execute this every once in a while and as soon as you open this container for the first time. It may take some time to execute.

• ros_ws: This directory is the main ROS workspace. It is where you will do most of your development.

Editing with VSCode

If you want to use VSCode to do most of your development and don't want to bother with having to manually open the Docker container and whatnot all the time, I highly suggest adding this VSCode extension pack. More information about remote development in VSCode may be found here.