List Methods of Selenium.WebDriver.DriverExtensions.HasNetworkConditions package of Selenium Framework

Try:

Watir::Browser.new( :phantomjs,
    args: '--proxy=localhost:8181'
)

When using Capybara there is no guarantee that a page has fully loaded (whatever that means) when visit returns, and there is no guarantee any actions triggered by clicking an element have completed (or even started) when it returns. This is because Capybara knows nothing about the status of the page other than what the "browser" tells it is visible on the page. Using just Capybara the best you can do for timings is the time between you telling the browser to do something, and the time an element, expected to be on the page when "loaded", appears on the page. In your instance that would mean something like this

SIGNIN_BTN = ".//div/button[contains(.,'Sign In')]"    
VERIFY_BTN = ".//div/button[contains(.,'NEXTBUTTON')]"

t1 = Time.now    
visit "http://google.com"    
signin_btn = find(:xpath, SIGNIN_BTN) # wait until signin button is visible on page
t2 = Time.now   
pageloadtime1 = t2-t1  
puts pageloadtime1 

t3 = Time.now  
signin_btn.click 
verify_btn = find(:xpath, VERIFY_BTN ) # wait until the verify button is visible on page
t4 = Time.now  
pageloadtime2 = t4-t3
puts pageloadtime2 

t5 = Time.now   
verify_btn.click
something = find(:xpath, './/xpath of something visible on page when action has completed')  
t6 = Time.now  
pageloadtime3 = t6-t5
puts pageloadtime3

Note, these measurements will all include some overhead since the waiting behavior while attempting to find the elements has overhead and 50ms delay between each attempt. Also note the SIGNIN_BTN and VERIFY_BTN xpaths now start with .// - You should get used to starting all xpaths used with Capybara with .// rather than // unless you really need // and understand the difference (it breaks scoping on the page) - https://github.com/teamcapybara/capybara#beware-the-xpath--trap

A proper way to do it for all js tests is to add following inside spec_helper.rb RSpec.configure block

config.before(:each, js: true) do
  Capybara.page.driver.browser.manage.window.maximize
end

to maximize the window. Change to resize_to(x,y) to set any window size.

EDIT: If you happen to be using Poltergeist the correct way to do it is

config.before(:each, js: true) do
  Capybara.page.driver.browser.resize(x,y)
end

From MDN's URL documentation:

It is worth noting that the part after the #, also known as the fragment identifier, is never sent to the server with the request.

Check your server logs; you'll see that any request to /widget-web/#/ is sent as a request to /widget-web/. The # prevents anything after it from being sent to the server.

What is a URL? docs

Try the following,

driver.switch_to.frame "ext-comp-1465"

or

driver.switch_to.frame driver.find_element(id: 'ext-comp-1465')

or

driver.switch_to.frame driver.find_element(name: 'ext-comp-1465')

or

driver.switch_to.frame driver.find_element(xpath: '//iframe[starts-with(@id,"ext-comp-"] ')

 require 'watir'
 b = Watir::Browser.start 'http://www.gmail.com'
 t = b.text_field id: 'entry_1000000'
 t.exists?
 t.set 'your name'
 t.value

put exact id of that text field . id or xpath or name etc

The following two should work:

expect(page).to have_link('', href: 'http://project/guides/basics/')
expect(page).to have_selector("a[href='http://project/guides/basics/']")

I was really hoping there is a better looking way, but unfortunately, I can't find one.

Javascript may interfere. The page had some unnecessary dependencies to Squeezebox, Mootools-1.2.5 and Function.cbb.js, so I just removed them and refactored to use more css functionality (like hover) and specialized js. I'm not impressed with this. Nonetheless, I hope this answer would help others.

update: Seems like Mootools is incompatible with Selenium. Without including any other JS library/code than the mootools core library, it breaks. I've tried both 1.2.5-core and 1.3.2-full-compact.

The following two should work:

expect(page).to have_link('', href: 'http://project/guides/basics/')
expect(page).to have_selector("a[href='http://project/guides/basics/']")

I was really hoping there is a better looking way, but unfortunately, I can't find one.