Google Chrome

Locator recording in windows fails with Google Chrome

When you are testing a Web application in Google Chrome, locator recording in windows fails when multiple windows are open during application configuration in the Google Chrome instance, in which the application is running. If you close the other Google Chrome windows during application configuration, the error will not appear.

Background applications in Google Chrome prevent automation support from loading

When you want to test a Web application with Google Chrome and the Continue running background apps when Google Chrome is closed check box is checked, Silk Test cannot restart Google Chrome to load the automation support.

Silk Test cannot record locators in modal dialogs when Windows Aero is disabled

If Windows Aero functionality is disabled, modal dialogs are not recognized and locators in such dialogs cannot be selected. As a workaround, use the Locator Spy or the Identify Object dialog box to manually create and verify locators while a modal dialog is displayed.

Silk Test does not display embedded PDFs

With Google Chrome 42 or later, Google Chrome by default blocks the NPAPI plug-in, which is used to display embedded PDFs. Because of this, Silk Test does not display embedded PDFs in Google Chrome 42 or later, but instead downloads the embedded PDFs.

  • If you are using Google Chrome 44 or prior, you can unblock the NPAPI plug-in in Google Chrome, by typing the following into the address bar:
    chrome://flags/#enable-npapi
  • If you are using Google Chrome 45 or later, the NPAPI plug-in is completely removed from Google Chrome, without an option to re-enable it, and all PDFs are downloaded.

Connection timeouts when executing tests against Google Chrome 49 or prior

When executing tests against Google Chrome 49 or prior on a slow machine, you might experience connection timeouts, causing the tests to fail. The following error message is displayed:
Error executing '*'. Communication with browser automation timed out.

To avoid such connection timeouts, ensure that the test machine has enough processing power. For example, if you are testing on a slow virtual machine (VM), you could enhance the processing power by adding an additional CPU core to the VM.

Setting the UserDataDir through the registry breaks the Google Chrome support when using Google Chrome 66 or later

If the user data directory is set as a policy in the registry through the key HKEY_LOCAL_MACHINE\Software\Policies\Google\Chrome\UserDataDir or the key HKEY_CURRENT_USER\Software\Policies\Google\Chrome\UserDataDir, and you are testing a web application on Google Chrome 66 or later, the base state might fail with the following error message: Failed to start application 'GoogleChrome'. unknown error: DevToolsActivePort file doesn't exist . This is a known issue in ChromeDriver: https://bugs.chromium.org/p/chromedriver/issues/detail?id=2513.

As a workaround, perform one of the following:
  • Remove the registry key.
    Note: This is the only way to enable parallel testing with Google Chrome if this issue occurs.
  • Set the user data directory in Google Chrome to the same directory as in the registry key.
    1. Open the Edit Browser Application Configuration dialog.
    2. Select Google Chrome as the browser type.
    3. Set the user data directory in the Connection String field: goog:chromeOptions={"args":["--user-data-dir=<user data directory>"]}. For example, if the value in the registry is C:/temp/chromeUserData, type goog:chromeOptions={"args":["--user-data-dir=C:/temp/chromeUserData"]}.
    Note: Parallel testing with Google Chrome will not work, even if you apply this workaround.