Running Tests on Multiple Operating Systems - Basic Steps

To check that your application works correctly on different Windows versions, you need to test it on all of them.

Creating tests for different Windows versions may be tricky, because there are some differences between them. For example:

• Windows system paths and file names: C:\Users vs C:\Documents and Settings, Documents vs My Documents, and so on.

• Different UI in standard system dialogs (Open File and others).

• Different settings of User Account Control (UAC), which prevents applications from performing privileged operations unless they have administrator permissions.

TestComplete includes features that make it easier to create tests for different Windows versions:

• The Sys.OSInfo and aqEnvironment objects let you check the Windows version to include version-specific operations in your tests.

• The OpenFile and SaveFile methods let you work with the Open File and Save File dialogs in the same way on different Windows versions.

For common tasks and solutions, see the topics of the Running Tests on Multiple Operating Systems - Typical Tasks section.

The overall procedure of creating tests to be run on multiple operating systems contains the following steps:

1. Plan steps your test will perform and determine which steps will be common for all operating systems and which steps will be specific to particular operating systems.

2. Implement test steps. It is recommended that test steps be implemented as separate script routines, keyword tests, low-level procedures, and so on.

3. Group created steps so that steps specific to operating systems are run only on corresponding operating systems and common steps are run on all operating systems. To manage created steps, you can use test items or you can manage them from other tests. For more information on that, see Running a Test by Name.

4. Select the environment where your tests will be run. You can either use physical computers, or virtual computers with desired operating systems installed. Make sure you have network access to computers where the tests will be run and those computers have TestComplete (or TestExecute) installed.

5. Select an approach you will use to control test runs on other operating systems:

   • You can control the tests manually, that is, you have to log in to computers with the desired operating system running, copy the tests there and initiate a test run.

After you run your tests on several operating systems, you may want to gather results generated for each operating system, compare and analyze them. For each test run TestComplete generates a test log that contains information on all actions performed during the run.

Explore the test log generated for your tested application on different operating systems to learn whether the tests are successful. To view the test log generated for your tested applications on multiple operating systems, you can either log in to the remote computers where the tests were run and access the logs manually or you can use various features provided by TestComplete for managing logs.

For more information on working with TestComplete logs, see Test Results.