Just last week, the WebDriver specification officially became a W3C Recommendation, defining a standard way for web developers and browser engineers to automate the browser. This is a major step forward for web site and web app testing, as well as cross-browser interoperability initiatives like web-platform-tests.Over the past few months we’ve been working to implement the updated W3C dialect for WebDriver in Microsoft Edge—a generational improvement to automated testing on the web. Today’s Windows Insider Preview release (17692) includes our updated implementation, as well as making it a Feature on Demand, so it’s easier than ever to get started.
WebDriver needs to match the version of Microsoft Edge you’re testing against, which has historically required manually matching a standalone download of WebDriver to the appropriate version of Windows on your device.
Beginning with today’s preview release, we’ve made WebDriver a Windows Feature on Demand (FoD), which ensures that it’s always up to date automatically, and enables some new ways to get Microsoft WebDriver.
The simplest way to get started is simply to enable Developer Mode. Simply open the Settings app and go to “Update & Security,” “For developers,” and select “Developer Mode.” The appropriate version of WebDriver will be automatically installed.
You can also install a standalone version of WebDriver in one of two ways:
Search “Manage optional features” from Start, then select “Add a Feature,” “WebDriver.”
Install via DISM by running the following command in an elevated command prompt: DISM.exe /Online /Add-Capability /CapabilityName:Microsoft.WebDriver~~~~0.0.1.0
This also means that we will no longer be providing standalone downloads for Microsoft WebDriver going forward, however we will keep previous releases (RS4 and down level) available on our download page.
Actions API and new commands
The Actions API allows for low level input into the browser via action sequences, allowing developers to send multiple streams of input to test complex scenarios. Our implementation currently supports both mouse and keyboard input.
We’ve also added support for new commands including Get Timeouts, Get Window Rect, Set Window Rect and Get Element Property.
We’ve also implemented new logic for a number of commands, in order to improve interoperability and reduce test flakiness when running in Microsoft Edge:
Supporting CSS pixels for Set Window Rect, so that we scale properly for high-DPI devices when resizing
Adding calculations for in-view center point to impact what is/isn’t clickable and requires scrolling
Adding proper support for implicit wait to commands that were missing it
Using the Selenium implementation for Get Element Text
Testing PWAs and WebViews
These updates also apply to automated testing of PWAs and WebViews. So if you’ve been using WebDriver to test your web app you should now be able to enjoy these new benefits and bug fixes. We’ve also enabled WebDriver support for out of process WebViews.
As we move forward, we are working our way through the WebDriver web platform tests, fixing failures, and making sure our implementation is up to spec. As of our latest run we’re now passing 783 web platform tests out of 951. We’re tracking most of the remaining failures being as interoperability bugs or as missing features, and look forward to continuing to close the gap in future releases.
This is the most significant update since we first brought automated testing to Microsoft Edge with WebDriver. With these changes, it’s easier than ever to build interoperable web sites through cross-browser testing.
We encourage you to get started with the implementation in Windows Insider Preview build 17692 or higher, and share your feedback on Twitter or in the Feedback Hub app on Windows. Let us know what you think!
– Clay Martin, Program Manager, Microsoft Edge
Updated June 14, 2018 10:12 am