Per this discussion, it seems widget tests are under-documented in these ways:

• unclear why UI appears small with black regions above and below it
• unclear whether that behavior can be overridden or modified in any way
• it was unclear to me that driver and widget tests were a different thing, and how they differ. That said, this documentation seems pretty good. Perhaps I missed that when I was trying this stuff out.

Ultimately, what I want to do is use automate the execution of widget tests on my CI to take snapshots of the app at various points. This way, I can set up a feed for my test team to monitor to see if anything visual breaks during the development of an application. Of course, I will need to choose a resolution at which to take the snapshots, but I am unclear on how I can achieve this.

This should include:

• Add an example to test Landscape orientation in a widget test #2005
• Explain when to write a driver test and when to write a widget test #3609
• Add golden testing to 'An introduction to widget testing' page #4855
• Explain where test file should be stored on 'An introduction to widget testing' page #8022