When an automated test fails, it can sometimes be down to environmental issues or network instability and sometimes a rerun of the test would cause it to pass. It would be good if there was an option to rerun test x number of times on failure. This is particularly useful for automated deployments which run and can be hampered by tests that fail first time but potentially will pass a second or third time if run.
I find it frustrating that the scheduling server does not have a state or status for a scheduled test that was not run. In my experience if a test is scheduled and does not run it simply disappears from the results view. This happens for example if a test executor gets turned off and is not available to run the scheduled test. I would expect proactive notification from the scheduling service that a test was not able to be ran. Or keep the scheduled test in the results but have it red, with useful failure details in the log.
Can you please enhance the scheduling server/test executor to prevent compile time collision. I've run into a situation where I have a test executor that runs tests frequently, so it is compiling the pages.g.cs file each time. Well I sometimes go in and request for a test to be run remotely against that same executor - expecting that it knows when its busy and will get to my request when it can. ...and boom. The requests crash into each other and the executor gets a compile error in the pages.g.cs file. Frustrating. You should see about enhancing this so that the executor taking commands from multiple points (the scheduler, or the developer's box) don't collide.
When I restart the WPF application with the following code: ActiveApplication.Quit(); var runningApp = System.Diagnostics.Process.GetProcesses().Where(p => p.ProcessName == "WPFCalculator2"); Manager.LaunchNewApplication(@"D:\calculator\csharp\bin\Debug\AvalonCalculator2_WPF40Copy.exe"); Manager.ConnectToApplication(runningApp.FirstOrDefault()); the recorded is not able to attach again with RunToHere option and TS throws the following exception: Outer Exception Type: System.ArgumentException Message: Process with an Id of 13348 is not running. HRESULT: 0x80070057 (Official ID (if app.) = E_INVALIDARG, Error Bit = FAILED, Facility = FACILITY_WIN32, Code = ERROR_INVALID_PARAMETER) Call Stack: at System.Diagnostics.Process.GetProcessById(Int32 processId, String machineName) at System.Diagnostics.Process.GetProcessById(Int32 processId) at Telerik.TestStudio.Wpf.WpfRecorderWindowManager.AttachRecorderToApplication(Int32 pid, Action`2 callback) at ArtOfTest.WebAii.Design.DesignTimeExecution._runner_TestCompleted(Object sender, TestCompletedEventArgs e) at ArtOfTest.WebAii.Design.Execution.RunnerController.OnTestCompleted(ExecutionReturnValues result) at ArtOfTest.WebAii.Design.Execution.RunnerController.CallBackListener() If we just execute the test it works smoothly and executes all the steps after the restarting.
What I want to achieve is to run a combination of tests for different browsers. The request is based on the situation that one particular test isn't able to run in e.g. Internet Explorer and I want that test to run in Chrome. The example I have is: 001 (IE) 002 (Chrome) 003 (IE) 004 (IE) It would be nice if you could specify at each test in the test list at what browser it should run.
Currently when you move or delete the last coded step from a test the code-behind file still remains in the project. It would be nice if we delete the code-behind when no coded steps are left in the test. We can also put a confirm dialog, asking for permission to proceed, otherwise, some code-behind files can be deleted by mistake.
Based on the documentation, system path is a manual step http://docs.telerik.com/teststudio/features/test-runners/add-path-environment-variables It would be very beneficial when you are setting up multiple test machines to add system path automatically(or an option) as part of TestStudio MSI installation.
Even for a small sample of performance results, the load time is painfully long. Can you please load the History Tab values (total test times, over time) faster without needing to query the complete detailed result set? Possibly the query can continue loading in the background, after first displaying the high-level data. Thanks
Invoke OnBlur fails with the following exception in IE 11: ExecuteCommand failed! InError set by the client. Client Error: System.ArgumentOutOfRangeException: Index and length must refer to a location within the string. Parameter name: length at System.String.Substring(Int32 startIndex, Int32 length) at ArtOfTest.InternetExplorer.IECommandProcessor.InvokeEvent(IHTMLElement target, BrowserCommand command) at ArtOfTest.InternetExplorer.IECommandProcessor.ProcessActionCommands(BrowserCommand request) at ArtOfTest.InternetExplorer.IECommandProcessor.ProcessCommandInternal(WebBrowserClass ieInstance, BrowserCommand request, IHTMLDocument2 document) BrowserCommand (Type:'Action',Info:'NotSet',Action:'InvokeEvent',Target:'ElementId (tagName: 'input',occurrenceIndex: '74')',Data:'onblur--@@--null',ClientId:'Client_007162c7-bb26-4f8e-a68b-32c177fccba3',HasFrames:'False',FramesInfo:'',TargetFrameIndex:'0',InError:'True',Response:'System.ArgumentOutOfRangeException: Index and length must refer to a location within the string. Parameter name: length at System.String.Substring(Int32 startIndex, Int32 length) at ArtOfTest.InternetExplorer.IECommandProcessor.InvokeEvent(IHTMLElement target, BrowserCommand command) at ArtOfTest.InternetExplorer.IECommandProcessor.ProcessActionCommands(BrowserCommand request) at ArtOfTest.InternetExplorer.IECommandProcessor.ProcessCommandInternal(WebBrowserClass ieInstance, BrowserCommand request, IHTMLDocument2 document)') InnerException: none.
Schedule test list: Email settings Persist for the project. Currently you must fill out the attachment type, who the email is sent to, title, ect. Currently If you close out of the project these are lost. If you are in a project and do not close the project, it retains the settings. this should persist through the project.
I would like to make the list of agents "select all" checkbox, as checking multiple machines is tedious and if you misclick you have to start over as they are all deselected. I would also like to see a static set and save the "email to" in the results settings. I would also like to know who the test case owner is in the email as a option. I would like to have the ability to have the tests that fail in the list create a new test list, consisting of the failed test lists to run at a later time.
If scheduler is overloaded or stuck it will stop working. It should either attempt to auto-recover or send an alert that is not running.
Steps to reproduce: 1. Open a Test Studio project in Windows file explorer. 2. Copy and paste a test in a new folder within the projct. 3. Press the Refresh button in Test Studio so the test appears. Actual: The new test has the same ID as the original test. Expected. Test Studio should change the ID. This works if you select Add existing test from the project explorer.
There is no option to save only one test in the project. When you press the save button Test Studio saves all tests that are marked as dirty. The workaround is to go to the Recent project and reopen the same project you are working on. In this case a dialog with all dirty tests is displayed and the one can uncheck some of them.
It would be very beneficial to be able to select multiple test list results and be able to mass export them all to a single location as excel spreadsheets. Currently, it is very tedious to go through and manually export each test to excel, especially when you have a lot of test lists that get run as part of your suite.
Once tests are run using ArtOfRunner in command line, I like to extract the results (number of tests, passes tests, etc) so that I can send email with the pass rate and number of tests run. I can do this opening the results file and making into word/excel. There it gives all details. These details not saved in this file, SanityChrome 131250182933109322.aiiresult. Where can i see the tests passed out of line ? thanks, Sri
This feature will help teams sharing projects avoid issues with paths that include username.
Wondering if it's feasible to offer official support for electron-based apps at some point in the near term. Since the framework already utilizes front-end technology, is this something that we may be able to incorporate into a future release?