I'm unable to replicate behavior I see during a testcomplete test by hand
Heya - so, the purpose of the test I'm developing is to test our license manager. The expected behavior is that when our application starts, it ping's the license manager (LM) and the LM 'checks out' a license from our network hasp key, and holds it while the application runs, refreshing it very 10 minutes or so, until the application closes, at which point, the LM 'checks it in' to the hasp key and releases it for someone else to 'check out'.
When I run the test, and shut down the application (using .close(), .terminate() and actually replicating clicking the file-exit button), using testcomplete, the LM isn't isn't checking the license back in to the hasp key. Eventually, it will time out (when the hasp manager pings the LM to ask if it still needs the key), but it essentially leaves the license 'hung' for about 10 minutes. Obviously, a bad thing.
BUT
When I replicate the EXACT same test by hand (which is to say check the number of seats checked out, start the application, verify that one was added, then shut down our application and verify that it was checked back in) I cannot replicate the 'not checking it back in' behavior. The LM checks it back in every single time.
Any ideas?