Visual Regression Tests from Windows 10 11 (Ubuntu VM)
This page is if for someone who does most of their development on Windows and doesn't run OSX or Linux natively. It is probably the easiest way to run the visual regression tests if you are such a person.
You will need to create a Linux box in a virtual machine. You can run the virtual machine in either VMWare or Oracle VirtualBox as a host. This guide shows VirtualBox, but VMWare should work similarly.
Before running VirtualBox, you should enable 64-bit virtualization. If you are already running virtual machines on your computer you can skip this step. Depending on how your system was set up, these steps may already be done.
Boot your computer and enter the BIOS menu. In my case, this is done by holding down the delete key before the windows system boots (ASUS). Your system may be a little different depending on your BIOS vendor.
Virtualization is enabled in menu that looks like this (should be Enabled):
After your system boots, in Windows control panel, choose Programs/Turn windows features on or off. Make sure Hypervisor is off (box is unchecked):
Install VirtualBox (it's free). Find a good Linux server image. I used a Ubuntu image from OSBoxes.org. Download a server image, it will be about 2GB.
Then add the media from your image to VirtualBox:
Now create a new virtual machine:
Set the memory size to be about 1/2 of your physical memory.
Choose the image you just downloaded and added to your media list as the drive:
Then click 'Create'. In the settings for the machine, you should set your processor count to how ever many physical processors on your computer (e.g. 4 for an Intel i5). This will greatly enhance performance.
Then start the machine. It will boot up and dump you into a prompt, where you can login.
You only need a couple of tools to run vexflow. GIT:
sudo apt install git-all
And Node/NPM
curl -fsSL https://deb.nodesource.com/setup_12.x | sudo -E bash -
sudo apt-get install -y nodejs
Important note: I was unable to get vexflow to work with the latest version of Node(16.x at the time of this writing). The issue is with an NPM package called canvas that has a number of issues. There may be other solutions, but I found it easiest just to install the older version of node as shown above.
> npm -version
6.14.6
> node --version
12.18.3
You can use NVM if you have multiple node versions to support.
From here, you can follow the instructions for build and visual regression tests.