I'm not sure how I can phrase this, but it legit would work one day, then if I open it up the next it may not work. It happens once in awhile, but its enough where its starting to get annoying. The logs usually say things like there was an interruption during the test. Like just general things, and like I easily fix this sometimes but just closing the testing scenario and reopening it.
Use the same philosophy you use to find bugs in your application. ie, A good tester is curious of what's going on and dig deep until find the real culprit like a detective.
one things to investigate is timing TC is faster than most application so when TC is looking for object it may have not created,
sometimes it is there.
to guide you to right direction we need more information.
I have added delays in spots that is required. But I'm trying to explain to you, that there was nothing different. The only thing is really different is, I am opening test complete on a different day. I complete these tests, then save them. Re-open them another day to make sure they are still working and turns out some don't. The things it stopped at during a test, is as simple as clicking a program buttons then mousing over the menu to select a option. Like a drop down menu. I honestly, not even sure how can I "dig deeper". Its legit, it was working, and now its not. Is there some kind of logging or something I can provide you?
No errors or anything in the log other than "test run interrupted"? That's a bit odd... If there are other errors in the log, knowing what those are would be helpful in figuring things out. Anything like "Object not found" or things like that? Do you have the "Stop On Error" option checked in Tools -> Current Project Properties -> Playback? If so, as soon as you write an error to the log, it will stop execution.
So... guess if you could provide screenshots of the logs that are showing this behavior that would help debug further. As NisHera mentioned, it does sound, a bit, like a timing issue... but it also sounds like object identification issues. If you have your objects mapped using properties that may dynamically change from run to run, you will encounter any number of errors that, if that "Stop on error" option is checked, will halt your test run. So, it MIGHT be that you will need to rework some of your mapping criteria to be a bit more robust.
But, again, the logs will tell more... please provide whatever you can.
