Skip to content

Latest commit

 

History

History
131 lines (69 loc) · 10.1 KB

windows_readme.md

File metadata and controls

131 lines (69 loc) · 10.1 KB
Raw

Table of Contents

WoZ4U Guide for Windows

In this guide, we provide the steps and commands to install and run WoZ4U on Windows machines. These are slightly different from the process to install and run WoZ4U on MacOs or Linux, because:

  • Bash is not available per default
  • Python's virtual environment works slightly different
  • Environment variables are handled completely different

Thus, we found a dedicated guide to install WoZ4U on Windows beneficial, of not necessary.

Download the required frameworks

To begin, you will have to download and install a few additional software components:

  1. Install Git for Windows from here. The setup is quite long, but you should be fine to just leave all options on the default values. Alternatively, here is a guide explaining the individual steps in greater detail. After you have completed this setup, Git Bash is available on your system and you should be able to launch it in the start menu. You will need this to a) clone WoZ4U from GitHub, but also to type the commands to install the requirements and ultimately to start the main interface.
  2. Install Python 2.7 from here. Independent of your system architecture, you have to install the 32-bit version. This is because the NAOqi-API has only been compiled for 32 bit systems. Thus, if you install the 64-bit version Python 2.7 the NAOqi-API will throw errors and prevent you from using WoZ4U. Luckily, Windows 10 handles this well, and you can just install the 32-bit version of Python, even if you are on a 64-bit machine.
  3. Install PyCharm from here. This step is somewhat optional, however, we found that using Pycharm to set up a virtual environment is, by far, the most convenient and easiest way to do it. You can either install the free community edition, request an EDU license, or, if you have one, use your professional license.
  4. Download the NAOqi-SDK from here. Notice, how the SDK is only available in 32 bit. This is why you had to install the 32-bit version of Python 2.7 earlier. Extract the zip to any location, remember this location, you have to specify it in the next step.

Download WoZ4U

Download WoZ4U from GitHub using GitBash, which you installed earlier:

  1. Start the GitBash terminal by searching for it in the start menu. This will spawn a Bash terminal, looking similar to this:
  2. Using the cd command in GitBash, navigate to the location where you would like to install WoZ4U, for example on the Desktop: cd Desktop
  3. Download (clone) the WoZ4U repository onto your machine: git clone https://github.com/frietz58/WoZ4U.git. Don't paste this command (or anything for that matter) into your GitBash terminal using CTRL + V! Instead right-click into the terminal and select paste or type the command manually. If you press CTRL + V in GitBash, you will get this error.

You should now have a folder called WoZ4U at the location where you executed the git clone command. You can verify this by inspecting the location with the Windows Explorer.

Setting up the environment

This is the trickiest part of the installation, because it is very easy to mess things up, given how Windows handles environment variables. Thus, instead of doing this step in the terminal, we use PyCharm to do the heavy lifting and setting up of the environment. If you feel confident enough and know what you are doing, you can of course create the virtual environment via terminal, otherwise, just follow the steps below:

Pass NAOqi-API reference

For Python to know where the NAOqi-API is available, you need to add the location of the folder to the PYTHONPATH. This environment variable contains all the "locations" the Python interpreter will search for packages. Edit the script called set_paths.sh in the WoZ4U folder, for example by right-clicking it and opening it Notepad. If you double-click the file, Windows will treat it as an executable (because it is one) and run the file, but won't let you edit it.

In the code of the file, identify the first occurrence of the string "PYTHONPATH" (line 4). Adjust the value of in quotes so that it contains the path to the lib folder inside the NAOqi-SDK folder, that you downloaded and extracted earlier:

# PYTHONPATH="C:\ABSOLUTE\PATH\TO\LIB\FOLDER\IN\NAOQI-API-FOLDER";
PYTHONPATH="C:\Users\Finn\Desktop\WoZ4U\pynaoqi-python2.7-2.5.7.1-win32-vs2013\lib";

Create virtual environment

  1. Start PyCharm.

  2. Open the WoZ4U folder in Pycharm (File --> Open). PyCharm might work for a short moment because it scans all the files in the selected directory, but after that, you should see the files in the WoZ4U folder displayed on the left-hand side of PyCharm.

  3. Open the "Project settings" of the WoZ4U project: File --> Settings. Alternatively, press CTRL + ALT + S.

  4. In the project settings, navigate to the Python Interpreter settings (marked with 1) in the following screenshot). Under the interpreter settings, open the dropdown menu and press the Show all... option (marked with 2) in the screenshot):

  5. In the list of all available python interpreters (if you never used Python before, there should only be one interpreter called Python 2.7), select the plus button to add a new one (1) in the screenshot). Then, create a New environment (2) in the screenshot). Leave the location as default, which will create the virtual environment inside the project folder This is important because we assume this location of the environment later. Make sure that you select your Python 2.7 installation as the base interpreter (3) in the screenshot). Again, if you just installed Python, there should only be one option. Name the virtual environment as you like (default is venv), but you can also call it woz4u_venv if you prefer a more descriptive name.

  6. Press OK. Pycharm will now create your virtual environment. After the process has finished, press OKin the remaining dialogues.

To verify that the virtual environment was created, either in with the Windows Explorer or with the file view in Pycharm, check that at the location specified in the setup, a new folder has been created with the following structure:

Notice that there is a script called activate in the venv/Scripts/ folder. We will call this later to activate the virtual environment.

Install requirements

Now that we have a virtual environment, the last step is to install the python requirements for WoZ4U.

For this, activate the virtual environment we just created. This is done by "sourcing" the activate script inside the virtual environment's Script folder:

  1. In GitBash, navigate to the folder where the virtual environment is located. If you followed the guide without deviation, it will be located inside the WoZ4U folder: cd path/to/WoZ4U-folder/
  2. Activate the virtual environment source venv/Scripts/activate If you named your virtual environment differently, adjust the command to reflect this: source <name-of-venv>/Scripts/activate.

This should output, in parenthesis, the name of the virtual environment you just activated. If you left the default value in the PyCharm dialogue, it will output (venv). Further, GitBash will display the name of the active environment after every command you execute, while that environment is active. See the annotations in the following screenshot:

  1. Install the requirements inside of the currently active environment (because we just activate the virtual environment for WoZ4U, there won't be any conflicting package versions): pip install -r requirements.txt

That's it. You are now able to start the interface and connect it to a Pepper robot.

Running WoZ4U

The main command to start the WoZ4U interface is to tell Python to execute the file server.py. However, two other requirements must be met. The required packages (that have been installed in the virtual environment) must be available, as well as the NAOqi-API so that Python can send commands to Pepper.

If you just completed the installation guide, your GitBash terminal should still have the virtual environment active. Nevertheless, we provide a command that executes all the steps, so that it will also work after restarting your computer or after opening a new GitBash terminal.

  1. Make sure the GitBash terminal's working directory is set to the WoZ4U folder (cd \Path\to\WoZ4U)

  2. Start the server by running: source venv/Scripts/activate && source set_paths.sh && python server.py You can see the three separate commands, combined via the && operator. First source venv/Scripts/activate activates the virtual environment. Then source set_paths.sh executes the script set_paths.sh so that Python knows where to find the NAOqi-API. Finally, python server.py starts the main interface.

    The output of that command should be similar to this:

  3. To access the interface, start any browser (we recommend FireFox), and type into the URL field: http://localhost:5000/

    You can now use WoZ4U in your browser:

Head back to the main README to learn how you can configure the UI elements of the interface and how to establish a connection between the interface in the browser and your Pepper robot.