How to use callback_thread method of Selenium.WebDriver Package

Your code looks like syntactically incorrect, You should try as :-

driver.find_element(:xpath, "//button[normalize-space(text())='Click me']")

or

driver.find_element(:xpath, "//button[contains(text(), 'Click me')]")

Note :- Your button text seems as contains extra white spaces. So for avoiding these white spaces you should use normalize-space() in your xPath

Hope it will help you..:)

I was able to figure it out after few more searches:

Started the hub (on local machine): java -jar selenium-server-standalone-2.43.1.jar -role hub

Started the node (on remote machine): java -jar selenium-server-standalone-2.43.1.jar -role node -hub http://ipofhub:4444/grid/register -Dwebdriver.chrome.driver="C:\Ruby21\bin\chromedriver.exe"

I had to specify the path of chromedriver while starting the server. I defined in the PATH on remote machine, but only the above method worked.

I have the following code in the test script that I executed from local machine: caps = Selenium::WebDriver::Remote::Capabilities.chrome @driver = Selenium::WebDriver.for(:remote, :url => "<http://localhost:4444/wd/hub>",:desired_capabilities => caps)

You can get Chrome to start in incognito by default.

There's a solution for Windows at the bottom of the link below... If you're not using Windows, you can Google how to achieve this for you OS =)

https://productforums.google.com/forum/#!topic/chrome/sYaZkNW8II4

I guess all of these tabs are part of the same web page? I.e., all in the same HTML?

If that is the case, driver.button class: ['btn btn-primary'] is going to stop when it finds the first instance in the HTML, but that isn't the button you are looking for every time (it's the button in the first tab, where your script worked as you expected).

The best options in my mind are

• find a way to uniquely identify the button in each tab (for example, use id instead of class if possible), or
• pull all the buttons into a collection and click the button using its collection index after you figure out which index aligns with each tab. For example,

button_collection = browser.buttons(:class, ['btn', 'btn-primary']) button_collection[2].click # Will click the 3rd button in the collection

Note: This answer only applies to using the Selenium driver with non-deadless chrome, and once you're inside the shadowDOM you can only use CSS selectors which means you can't use any of the Capybara methods that involve Capybaras built-in selectors ('fill_in', 'select', etc). This may also break with any release of Chrome or chromedriver.

Because of the way chromedriver/chrome works, it isn't returning from whatever action you did to open the print window until the print window is closed. It is technically possible to work around this by triggering the print window asynchronously. But then you run into the issue that most of the print window is inside shadow DOM elements which the WebDriver spec doesn't currently provide an API for. You can however work around that using evaluate_script.

Therefore to open and then close a print window would be something like this

print_window = page.window_opened_by do
  page.execute_script('window.setTimeout(function(){ window.print();}, 50)')
end

page.within_window(print_window) do
   print_app = page.find('print-preview-app')
   print_app_shadow = print_app.evaluate_script('this.shadowRoot')
   print_header = print_app_shadow.find('print-preview-header')
   print_header_shadow = print_header.evaluate_script('this.shadowRoot')
   print_header_shadow.find('.cancel-button').click
end

Obviously all those calls could be chained together, the intermediate elements are only spelled out to make it clearer what is being done.

If you want to tell the IE version during run time, you can use DesiredCapabilities.

  DesiredCapabilities ieCapabilities = null;
  ieCapabilities = DesiredCapabilities.internetExplorer();
  ieCapabilities.setBrowserName("internet explorer");
  ieCapabilities.setVersion("Version Number");
  driver = new InternetExplorerDriver(ieCapabilities);

For more info about DesiredCapabilities use this link http://code.google.com/p/selenium/wiki/DesiredCapabilities.

In the comments you said that i need both IE 8 and 9. Actually it is not possible, Windows currently supports to install only one IE version in a box. The IEDriver used the installed version of IE to launch.

If you want to use multiple version of IE to test then the better option to go with Windows Virtual Machines. You can talk with virtual machines by using the RemoteWebdriver instances.

You need to require it

It should be something like [As you told both are in same folder]

require 'base_helper.rb'

class FindADealerPage < BaseHelper
 #PageObjects
  FIND_A_DEALER_SEARCH_FIELD =    {id: "abcd"}
  MAP_TAB   =                     {css: 'abcd abcd abcd'}
  SERVICE_CHECK_BOX =             {class: "abcd"}
  SUBMIT_SEARCH =                 {class: "abcd"}
  DEALER_RESULTS_FIRST_DEALER =   {css: "abcd"}

  def initialize(browser, wait)
   @browser = browser
   @wait = wait
  end
 ...
end

  Capybara.register_driver :selenium do |app|
    Capybara.app_host = "http://google.com"
    Selenium::WebDriver::Firefox::Binary.path = ENV['CUSTOM_FF_PATH'] if ENV['CUSTOM_FF_PATH'].present?
    profile = Selenium::WebDriver::Firefox::Profile.new
    profile.assume_untrusted_certificate_issuer = ENV['SKIP_CERT_ISSUER'].present?
    profile['browser.download.dir'] = "#{Rails.root}/tmp/webdriver-downloads"
    profile['browser.download.folderList'] = 2 # implies custom location
    profile['browser.helperApps.neverAsk.saveToDisk'] = "text/csv,application/pdf,application/doc,application/docx,image/jpeg"
    profile.native_events = true
    Capybara::Selenium::Driver.new(app, :browser => :firefox ,:profile => profile)
       #Capybara::Selenium::Driver.new app, :profile => profile
  end

Try with this configuration

This is a known issue in chromedriver. As mentioned in the github

This is a long standing issue with Chromedriver. The window resizing doesn't work with the chromedriver and the community is silent on this bug fix. If the resizing is not critical for your tests, I would suggest you avoid the manage() method altogether and use a workaround of having your browser window always maximized using the ChromeOptions object at the launch of your chromedriver instantiation.

Though it is not happening to everyone, the issue still exists. It is mostly resolved by updating chrome or chromedriver to the latest version(still wont work for most people)

So better not to use maximize or resizing activity on chrome .However if you want, you can go with ChromeOptions which is suggested by ChromeDriver

ChromeOptions options = new ChromeOptions();
options.addArguments("start-maximized");