Forum Discussion
Hassan_Ballan
To answer your questions.
1. DPI is ok, cmd prompt returned 96.
2. Can't really check because its a remote desktop witch adjust to the screen size available, a little less than 1920 x1080. My Jenkins pipeline also had the same issue and it uses a fixed screen size of 1920 x 1080.
3. Cleaned the %LocalAppData%\SmartBear\ folder. It helps, but only for the first run, all runs after that have the same issue again.
[TestComplete] Test runner exit code: 2.
[TestComplete] [WARNING] Error: There was an attempt to perform an action at a point, which is beyond the screen..
[TestComplete] [WARNING] Errors occurred during the test execution.The search box VisibleOnScreen property returns false, and search box size L,T,W,H = 2510,328,186,31 is supposedly of screen. Although the browser shows this correct.
Since DPI is confirmed at 96, this no longer looks like a scaling issue. The large X coordinate (~2500) and the fact that cleaning %LocalAppData%\SmartBear\ only fixes the first run suggest this may be a window state or coordinate synchronization issue after using Maximize(), especially in RDP or CI environments.
I would avoid BrowserWindow.Maximize() and instead explicitly restore and set the window size using Position(0, 0, 1920, 1080). In remote or VM sessions this is typically more stable than maximize.
As a workaround, since clearing %LocalAppData%\SmartBear\ helps on the first run, you could add a step in your CI pipeline to clear that folder before starting TestComplete.
If this started after moving to Chrome 145, you could also temporarily fall back to a previously supported and stable browser version to see if the behavior changes.
If the issue persists, I would recommend opening a support case with SmartBear so they can investigate a potential Chrome 145 + TestComplete 15.76 interaction.
Finally, if you are accessing a VM from a desktop with a single monitor setup, make sure both the host machine and the VM have matching DPI and resolution settings. In remote scenarios, mismatched scaling or display configurations between host and VM can still cause coordinate inconsistencies even when resolution appears correct.
Also, can you confirm whether this reproduces on a local non-VM desktop machine?
If this resolves your scenario, marking it as the solution helps future readers find it quickly.