It's making your test fail because an OverLappingWindow event that has an unmodified LogParams will log an error. What you need to do is modify the properties of LogParams to change what is automatically logged from an error to a warning or, even, completely bypass the logging.
Okay. Why is it different than UnexpectedWindow, and what would I do if I want it to behave the same way? How do I modify LogParams?
IIRC, Unexpected windows are windows that open up that are modal in nature... in other words, you can't click "past" it on the components behind. So... they are more like a pop-up coming from the application under test that is interfering with your test flow and you can't do anything without getting rid of it.
Overlapping Windows are more like, in the middle of the test, a MS Outlook notification pops up on screen and gets in the way of the test. You can, technically, minimize it or drag it out of the way if you wanted to if you were a human user and continue working.
So... subtle differences which is why there are two different handlers.
Concerning something you mentioned at the top in the OP... I understand the convenience of having TC stop on any error... but I've found that it usually works better if I dictate when to stop... that way these unexpected halts don't happen and I can better control how things flow.
