Today for our brown bag lunch session we watched the second half of a video by Brian Keller, following the first part here from last week.
Brian discusses many practical issues, including platform support (5:33), the coded UI Test skeleton (10:00), variables generated by recording (13:20), finding controls (16:15), times delays, how fast to run the test, what speed? (19:15).
Brian has a number of tips here on speed of running automated coded UI tests, including how to watch threads being freed up as an indicator of when to continue tests.
Using Feature Pack 2 editor, removing comments and steps (21:00), adding delays and splitting methods (24:00). How to continue on error (26:00), handy way to add extra steps in for assertion logic (28:30-29:00), test for fail (31:00). How to close failed apps (32:00), adding databinding to UI tests, to provide test data (35:00), how to take snapshots (38:00), image of an app (39:20), searching and filtering information (41:41), more on databinding (42:21), friendlier asserts (42:53), desktop capture (43:33), how to take Microsoft Test Manager tests and make into Coded UI Tests (43:44), from test scripts (46:00).
There’s a lot covered, but useful as a quick guide into how to start off automated UI tests on the right foot, and how to get out of the problems before they start.