Manual Tests Overview
A manual test is a text file or file compatible with Microsoft Office Word 2003 or later that describes the purpose of the test and includes a sequential list of steps for a tester to follow.
You can create a test in Word format only if you are using Microsoft Office Word 2003 or later.
Choose this test type when the test you need to run is difficult or impossible to automate, such as when the connection between the computer and its power supply is interrupted, or when you want to test for user interface details or behavior. Manual tests are the best choice when the steps of the test must be completed by a test engineer, not by an automated script.
Another use of a manual test could be to refine or examine your test environment at a specific point in the testing process. For example, you might use a test run that contains several automated tests such as unit tests. After the first unit test runs, but before the second unit test runs, you might want a tester to perform specific manual tasks or to manually record certain metrics. You can describe these tasks in a manual test and have the tester perform them at the correct location in the test run, namely between the two unit tests. This scenario would require that you create an ordered test, because using an ordered test is the only way to guarantee that your tests are run in the order you want.
You cannot run a manual test remotely, and you cannot run manual tests from a command line.
Using a Manual Test to Obtain Code Coverage Information
You can use a manual test to obtain code coverage information about binary files. First, you must instrument all the binary files for which you want to gather code coverage information. This means selecting the executable files or DLLs in the Code Coverage page of the test run configuration editor, as described in How to: Obtain Code Coverage Data.
Next, while running the manual test, the tester starts the executable file. Code coverage information is gathered on this executable file while it is in memory, if it has been instrumented. If it, in turn, loads any DLLs that have been instrumented, code coverage information is gathered on those DLLs also.
Finally, while running the manual test, the tester starts an executable. This executable may be an instrumented artifact, and it in turn may load other instrumented artifacts, such as DLLs. As instrumented artifacts are running, code coverage information is gathered for them.
It is strongly recommended that the tester stop the executable to remove it from memory before finishing the manual test. This helps ensure that all the code coverage data is gathered successfully.