Introducing our new test results view
Hello there! We’re super excited to show off our new test results view. Now you might be reasonably asking “Hold on! Test result view? I thought you already had this?”. And you would be correct - we already had a way to look at test results but it was something extremely basic and not really convenient to use.
As a reminder, here’s how it used to look:
Our original assumption was that everybody will consume test results via the Junit .xml file that we provide for every test run and then delegate the actual test result UI to their CI provider of choice. Turns out, we’re utterly wrong and there are lots of use cases where it makes sense to look at test results directly in emulator.wtf. With our old test result view, it took ages to actually find out which test cases failed and why. This negated all the test run speed advantages our core service has. Whoops.
But that’s all history now. Let me introduce you to our new test result view in all its glory:
We do believe that running your tests as fast as possible is extremely valuable but it’s not the full picture. An equally important part of test monitoring is the ability to quickly identify what’s failing and why.
We combine test cases from all shards and devices into a single list, allowing you to focus on what’s actually important - finding failing test cases as quickly as possible. Our new test results page surfaces the exact failure reason for a test case as the first thing on the screen, followed by a high-fps video recording (read how to enable video recording so you can visually confirm what happened during the test failure. It’s also possible to view and download full logcat logs for the test run itself and compare the same test across different device models.
We’ll continue improving and tweaking the test result experience, including surfacing test failures directly in your Git provider or CI service, showing test statistics and trends.
Check out an example of our new test result view and let us know if you spot anything we could improve even further.