Interacting with an Infragistics UltraOptionSet
I've been tasked with automating a legacy (desktop MFC/WPF) UI that includes an Infragistics UltraOptionSet. TestComplete is unable to see "inside" this control, and does not provide explicit support for it. It's possible to call the .NET `set_CheckedIndex` method through script routines, but that is less than ideal as it 1) Does not mimic how a user would actually select an item in the radio button group 2) May not trigger various UI events that update things like validation, visibility, etc. in other controls. What are my options here?
Critical Issues with TestComplete 15.68.8 x64 – OCR, Recording, and Browser Stability Failures
Over the past 7–10 days, we have encountered persistent and critical issues with TestComplete version 15.68.8 x64 that are severely impacting our ability to perform automated testing. These problems have rendered the tool nearly unusable in our daily QA workflows. Specifically, we are experiencing the following major issues: ❌ OCR functionality is not working – the tool fails to recognize on-screen elements, blocking all tests that rely on visual recognition. ❌ On-Screen Action is unresponsive – we are unable to interact with the interface to add actions directly from the screen. ❌ Recording does not capture any actions – mouse clicks, text input, and other events are not recorded. 🔁 Frequent need to restart TestComplete – the application must often be closed and reopened in order to function, and even then, inconsistently. 💥 TestComplete is terminating browser sessions unexpectedly – browser instances used during test execution are being shut down randomly, causing test failures and data loss. ⚠️ Impact on Workflows: Key regression and smoke test suites are blocked. QA automation is significantly hindered. Time is wasted, productivity is reduced, and there’s an increased risk of untested code reaching production. The tool is currently in a non-working state for reliable daily use. 🧩 Environment Details: TestComplete Version: 15.68.8 x64 Issue onset: Within the last 7–10 days OS: Windows 10 / 11 (please insert the exact version if needed) Standard restart and troubleshooting attempts have not resolved the issues. ✅ Request: We kindly ask for an urgent investigation and resolution. Please confirm whether these issues are known, whether a fix is being developed, and if any temporary workarounds are available. A quick response would be greatly appreciated to restore productivity.Object spy can't recognize spring boot app JTable
Recently our java application has been upgraded to spring boot app and all the testcases are failing after the upgrade. TC Object spy recognize only the part of the JTable content but previously it could highlight the correct JTable area. The first image shows how the object spy identify the JTable area in the older version of our app and the 2nd image shows how it shows in the upgraded version of the spring boot app. When executing Click action on the first row of the JTable, TC clicks on the column header. So TC recognize wrong area as the JTable. How can I fix this? Please help. JTable creation is in ClassA and previously the application runs by calling ClassA. but after the spring boot upgrade, they have created a new main class ClassB and they are calling the ClassB methods from ClassA. After upgrade, application runs by calling Class B but all the table creation are still in ClassA.
POWERBUILDER & TESTCOMPLETE
Just wondering is there a workaround with the MSAA (pbdw*) to accommodate Powerbuilder 2019 & 2022. We migrated to 2019 and I remapped all of the datawindows in TestComplete so it would be recognized during runtime. Again, we migrated to 2022 and I'm to the same issue. I know about the supported version... Supported PowerBuilder Versions With TestComplete, you can create and run tests for applications created with PowerBuilder Classic version 12.0, 12.5, 12.5.2 and 12.6. I saw other posts with similar questions regarding the same issue. Support ticket created Thanks 🙃
An increase in OCR API failures
Hi there, i'm wondering if anyone else has noticed a change in the performance of the OCR API? We run a test suite daily which likely has ~500 OCR API calls. Up until about a month ago, I'd only ever seen one OCR API failure, however we are now getting 3-4 a day with nothing else changing in regards to the test suite and hardware it's run on. The error is 'Failed to connect to https://ocr.api.dev.smartbear.com. A connection with the server could not be established.' I have raised with support, but they only really had assistance for if the OCR API wasn't working at all, which isn't the issue here, just the reliability. Does anyone have any advice on what could potentially cause an increase in these failures? (or any troubleshooting steps) Or, has anyone else noticed the same trend?
Qt6-Objects are not recognized
Hello, After Qt6.5-Upgrade of the software any Qt-Objects are not recognized by TestComplete (15.54, neuest at this moment) any more. The TestComplete Tests worked with the software with Qt6.2 or Qt5 without problem. Is there any way to make it work, other than waiting for TestComplete update? We are also wondering if we can find the Qt6.5(LTS)-Support plan somewhere. Thank you.
objects previously recognized on desktop app
When beginning this project, I went through the exercise of getting objects to be recognized by test complete. I have a number of objects whose class is reported as WindowsForms10.Window.8.app.0.198698e_r8_ad1 or some variation. I was able to get recognitions with: adding appropriate the class names to Project/Properties/Open Applications/MSAA. I then began happily writing my test and enjoying every happy minute. What changed, you asked? I don't know. I make no changes to code or project properties. Immediately before observing this behavior, I had accepted the TestComplete dialog to update to the new version. Of course that is it, thought I. So I downgraded to TC 15.54.3.7 (which is the only version I have around. Waiting on SmartBear to tell me why my TC licenses are not showing up in the license manager.) I am experiencing the same problem in this previous version of TC. The only other thing I can think of that changed is that the AUT's database was restored. But the behavior that I am seeing in TC began before this. And in no way did this change the user interface. I looked at installed extensions, they are all installed except HP something something which I don't have. Any ideas from anyone on what might be going on? ThanksTestExexute sometimes not recognizing object
When I execute my Desktop application using TestExecute, sometimes it would not recognize one of the objects. However, after rebooting the computer and restart the test again, the object could be recognized again. Test runs with TestComplete does not have this issue. How do I resolve this issue so that my tests can be reliably ran every time with TestExecute?Reset of the Test Execute enable extensions
Hello, We use Test Execute on our Azure Virtual Machines, and our test scripts rely on the legacy OCR module. Our team writes scripts using TestComplete version 15.54.x.x, with a similar version for TestExecute. Recently, we updated TestExecute to support testing web functionalities using the web module. We update to the version 15.71.4.11. However, after the update, we noticed that the legacy OCR module was disabled when running tests. We manually enabled the module in the extensions settings before running the tests: Unfortunately, the next day, some of our tests failed due to the OCR module being disabled again. It appears that the extension settings reverted to their default state, as if TestExecute had just been installed. We’re uncertain why this is happening, but we suspect it may be related to license availability. The TestExecute logs indicate that, at the time the OCR module failed, there was no available license for the machine: We believe the issue might occur when the machine attempts to execute tests while waiting for a license, and the connection is not successfully established. This could potentially cause the settings to reset to their default state. Our questions are: Why is this happening? What steps can we take to prevent this issue from occurring again? Thank you for your assistance. Best regards, Janusz
