Hello Support,

I'm running into this situation (System.NullReferenceException: Object reference not set to an instance of an object
) couple of times every day.

We have more than 50 different scripts; every other test has coded steps. All the scripts are working as expected.

But, while running scripts for some reason (say after 20 scripts) we closed browser forcefully it throws below exception. Now, we are using QA SP2; but I see same issue since long time. I’m expecting to see Failed -15 Passed Out of Total 20 executed.

Finally, we are ending up with Failed - 0 passed out of Total 0 executed.

Screenshot attached.

Is there any way to solve this issue?

Your help is really appreciated.

Thanks,

Madhu.

------------------------------------------------------------
'7/12/2011 5:11:35 PM' - Starting execution....
'7/12/2011 5:11:39 PM' - Detected a Silverlight Test. Setting EnableSilverlight=True
'7/12/2011 5:12:03 PM' - Failure detected during execution. Details:
------------------------------------------------------------
'7/12/2011 5:12:03 PM' - System.NullReferenceException: Object reference not set to an instance of an object.
   at ArtOfTest.WebAii.Design.Execution.ExecutionEngine.CanRunAgainstBrowser(Browser activeBrowser, AutomationStep step, TestResult activeResult)
   at ArtOfTest.WebAii.Design.Execution.ExecutionEngine.InternalExecuteSteps(AutomationStepList stepList, TestResult testResult, Object codeBehindInstance, Boolean& stopExecution)
   at ArtOfTest.WebAii.Design.Execution.ExecutionEngine.InternalExecuteTestIteration(Object codeBehindInstance)
   at ArtOfTest.WebAii.Design.Execution.ExecutionEngine.InternalExecuteTest(Test test, TestResult initializationResult)
   at ArtOfTest.WebAii.Design.Execution.TestExecuteProxy.ExecuteTest(Test test, Settings settings, String deploymentFolder, String binariesFolder, ExecutionType exeType, Guid firstStepGuid, Guid lastStepGuid, IntPtr existingBrowserHandle, DebuggerOptions debuggerOptions)
------------------------------------------------------------
'7/12/2011 5:12:03 PM' - Test completed!

I have a test to log in to our application.  I have another test that adds a new record into the app.  When I combine these 2 tests in a list, the first one logs in sucessfully, but then the second test then opens a new browser window.  This second window obviously doesn't have the app loaded and so I get a failure for the run.  I will get a successful run if I call the login script from within the second one, but my goal was to have a bunch of small tests that I could string together however I wanted.  I don't want to have to login, do one test, and log out.  Then log back in, do another test, and log out.
How do I make sequential test scripts recognize the browser instance that's currently open/running?

Thanks,
Steve

Question...

I have a search page that returns rows displayed on the web page by a default filter…  I want to verify the number of results on the page is the same number of results returned when I query the database.  How is this done using your tool?

On older versions of Tests Studio / Web UI, when using "run to here", it would put me in record mode after the step I was running to.
Now it wont record anything unless I go back and click, "record next step" or it will record the step in a random location in my test.
THis drives me nuts.
I like the way it was before.
I would click, "run to here", and then I could just start recording and the step(s) would be inserted after the step I ran to.

We have a good number of tests that we are looking to automate which require that the test uploads a .txt or .csv file to our system.  This is easy to accomplish on my laptop - but my question is what would be the best way to set this up if we plan on running these same tests on different workstations or using the Runtime Edition on a Rackspace server?  In other words, how we can ensure that the files to be uploaded can be found by the tests regardless of which machine the tests are being executed on?  Is there a way to add these files as attachments to the project or individual tests?

I performed the following steps:
1. Recorded a script.
2. Checked element identification logic to ensure that the identification logic did not include any dynamically created ID's & that elements were unique.
3. Ran the script.
4. Script failed because several elements that had been recorded could no longer be found on screen.
5. Used Find Element - Live Action to review element definitons and validate those definitons. The identification logic was fine, but validation failed
6. Opened the page where the elements were located, clicked on the elements to find out more info about them - changed identification information in Telerik.
7. Attempted to validate elements with new definitions - failed again.


Note: all of the elements that I am having this issue with are found on 1 frame.  I've tried identifying the item with the default frame inforamtion given to me by Telerik and I've tried changing the frame definition, but nothing I'm doing seems to be working.  Other frames are identified exactly the same as the problem frame and the system is having no problem identifying elements on those frames, so at this point I'm out of ideas.  Suggestions would be appreciated.

When I open a test, I see the end of the test not step 1.2.3.etc.
Also when I delete a step or group of steps, it throws me to the last step after the delete is done.
Then I have to find my way back to the steps I was working with.
3rdly, When I copy steps from one test to another, the order or the steps is reversed.
This does me no good at all.

I can record clicking on external links okay.
When I run the test after, I get an error on the connect to step.
I see some of the sites redirect so I copied the location it lands on and pasted it in to the connect step.
This fixed many of the connect steps but not all of them.
I still get errors on some even though I see everything happen correctly.
What can I do to get rid of the errors in the connect to steps?

After installing the Test Studio trial edition, tests created using the WebAii framework (free version) no longer run on the same box that Test Studio is installed. Our trial ended with a purchase of the commercial license, but we'll need to continue running tests on the existing framework until we can migrate completely.

The error I'm getting is:

Unit Test Adapter threw exception:
Could not load file or assembly 'ArtOfTest.Common, Version=2010.3.1307.0, Culture=neutral, PublicKeyToken=14a0200bfcbb7b62' or one of its dependencies. The system cannot find the file specified..

I don't quite understand what could be the problem. The file in error is still in the same place as it was before the trial install. 

I'm using
Visual Studio 2010
ArtofTest WebAii
Microsoft Test Tools
Silverlight UI testing

I have recorded a test using IE9 and it successfully replays in IE9 and Firefox 5. The same tests fails in both Chrome 12 and Safari 5 at the same point. In each case these other browsers are launched and show the app's login page.

The test is to click in the user name window, type the user name, tab to the pw window and type in the pw. The first step of the test is to navigate to the app at localhost:port#/index.html  which it does in all browsers. The point of failure is step 2 labelled: "radmaskedtextbox: click with '1' offset action" where it is to click in the user name window. It apparently times out? To ensure the environment is copacetic I ran the test in IE at the beginning and end of trying to run the others, and of course it succeeded both times.

I have installed as per directions the add-ins for Safari but don't know if Chrome needs one.

Lee