We Keep Coding

Python Selenium and Chromedriver CentOS8

I have a python script which use selenium and chromedriver.
It runs on my CentOS8 VPS perfectly for 3 days without any problem.
But since this morning, the script launched, wait almost 80 secondes and display this :
[12/Jan/2021 23:04:51] ERROR - Failed : Message: chrome not reachable
Traceback (most recent call last):
File "script.py", line 55, in <module>
driver = launch()
File "script.py", line 37, in launch
browser = webdriver.Chrome('/usr/bin/chromedriver',chrome_options=chrome_options)
File "/usr/local/lib/python3.6/site-packages/selenium/webdriver/chrome/webdriver.py", line 81, in __init__
desired_capabilities=desired_capabilities)
File "/usr/local/lib/python3.6/site-packages/selenium/webdriver/remote/webdriver.py", line 157, in __init__
self.start_session(capabilities, browser_profile)
File "/usr/local/lib/python3.6/site-packages/selenium/webdriver/remote/webdriver.py", line 252, in start_session
response = self.execute(Command.NEW_SESSION, parameters)
File "/usr/local/lib/python3.6/site-packages/selenium/webdriver/remote/webdriver.py", line 321, in execute
self.error_handler.check_response(response)
File "/usr/local/lib/python3.6/site-packages/selenium/webdriver/remote/errorhandler.py", line 242, in check_response
raise exception_class(message, screen, stacktrace)
selenium.common.exceptions.WebDriverException: Message: chrome not reachable
No modification have been made, why does it fail now ?
I don't have any screen on my VPS so I can't see more information.
Here is some info :
yum info on chromedriver :
Nom : chromedriver
Version : 87.0.4280.88
Publication : 1.el8
Architecture : x86_64
Taille : 27 M
Source : chromium-87.0.4280.88-1.el8.src.rpm
Dépôt : #System
Depuis le dé : epel
google-chrome --version :
Google Chrome 87.0.4280.141
Begin of the script :
from dotenv import load_dotenv
from logger import logger as l
from selenium import webdriver
from selenium.webdriver.common.by import By
from selenium.webdriver.common.action_chains import ActionChains
from selenium.webdriver.support import expected_conditions
from selenium.webdriver.support.wait import WebDriverWait
from selenium.webdriver.common.keys import Keys
from selenium.webdriver.common.desired_capabilities import DesiredCapabilities
from selenium.common.exceptions import NoSuchElementException
from selenium.webdriver.chrome.options import Options
import time
import sys
import subprocess
load_dotenv(verbose=True)
dotenv_path = '.env'
load_dotenv(dotenv_path)
def launch():
chrome_options = Options()
chrome_options.add_argument('--headless')
chrome_options.add_argument('--no-sandbox')
browser = webdriver.Chrome('/usr/bin/chromedriver',chrome_options=chrome_options)
l.info('Started Chrome')
return browser

Problem solved but don't understand how.
I just restart my VPS (reboot), and ... it works again.
Weird
EDIT : Find why !
I just made a mistake at the end of my script :
b.close();
But "b" don't exist, my driver variable name is "driver".
The exception was catched and not displayed, so I don't saw anything.
But today, I launch a "top" command, and see all "chrome" process running in background.
Probably after several days, memory was full, and Chrome can't launch.
The error was not clear but anyway, it was my fault.

Thumb rule
A common cause for Chrome to crash during startup is running Chrome as root user (administrator) on Linux. While it is possible to work around this issue by passing --no-sandbox flag when creating your WebDriver session, such a configuration is unsupported and highly discouraged. You need to configure your environment to run Chrome as a regular user instead.
Solution
Remove the following chrome_options:
--no-sandbox
and execute your code as a non root user.
Outro
Here is the link to the Sandbox story.

Selenium not using Default Chrome Profile

Selenium will not load my default Chrome Profile and I cannot figure out why. I have tried both Profile 1 and a Default profile with the same error (below). I have confirmed with Task Manager that all Chrome windows are shut down before running this code. Any thoughts?
from selenium import webdriver
from bs4 import BeautifulSoup
import pandas as pd
import numpy as np
from webdriver_manager.chrome import ChromeDriverManager
from selenium.webdriver.chrome.options import Options
import os
os.system("taskkill /f /im geckodriver.exe /T")
os.system("taskkill /f /im chromedriver.exe /T")
os.system("taskkill /f /im IEDriverServer.exe /T")
os.system("taskkill /f /im chrome.exe /T")
driver2 = r"C:\Users\xxx\.wdm\drivers\chromedriver\87.0.4280.20\win32\chromedriver.exe"
options = webdriver.ChromeOptions()
options.add_argument("user-data-dir=C:\\Users\xxx\\AppData\\Local\\Google\\Chrome\\User Data\\Profile 1")
driver = webdriver.Chrome(executable_path=driver2, chrome_options=options)
driver.get("https://www.google.co.in")
Traceback (most recent call last):
File "C:\Users\xxx\OneDrive\Python\pyReportRun.py", line 16, in <module>
driver = webdriver.Chrome(ChromeDriverManager().install(), options=options)
File "C:\Python38\lib\site-packages\selenium\webdriver\chrome\webdriver.py", line 76, in __init__
RemoteWebDriver.__init__(
File "C:\Python38\lib\site-packages\selenium\webdriver\remote\webdriver.py", line 157, in __init__
self.start_session(capabilities, browser_profile)
File "C:\Python38\lib\site-packages\selenium\webdriver\remote\webdriver.py", line 252, in start_session
response = self.execute(Command.NEW_SESSION, parameters)
File "C:\Python38\lib\site-packages\selenium\webdriver\remote\webdriver.py", line 321, in execute
self.error_handler.check_response(response)
File "C:\Python38\lib\site-packages\selenium\webdriver\remote\errorhandler.py", line 242, in check_response
raise exception_class(message, screen, stacktrace)
selenium.common.exceptions.WebDriverException: Message: unknown error: Could not remove old devtools port file. Perhaps the given user-data-dir at C:\Users\xxx\AppData\Local\Google\Chrome\User Data\Profile 1 is still attached to a running Chrome or Chromium process

chrom_options.add_argument("user-data-dir=C:\\Users\robert.car\\AppData\\Local\\Google\\Chrome\\User Data")
chrom_options.add_argument("profile-directory=Profile 1")
user-data-dir considers profile as default , and you don't have to specify that . If its something else specify it through profile-directory argument
Step to create a profile:
open : chrome://version in address bar
copy the user dir folder completely to eg c:\tmp\newdir
open the copied user data (newdir) and search for folder called Default . This is the profile folder.
rename the Default folder as "Profile 1"
Now to use this :
chrom_options.add_argument("user-data-dir=c:\\tmp\\newdir")
chrom_options.add_argument("profile-directory=Profile 1")

This error message...
selenium.common.exceptions.WebDriverException: Message: unknown error: Could not remove old devtools port file. Perhaps the given user-data-dir at C:\Users\xxx\AppData\Local\Google\Chrome\User Data\Profile 1 is still attached to a running Chrome or Chromium process
...implies that the ChromeDriver was unable to initiate/spawn a new Browsing Context i.e. Chrome Browser session as the user-data-dir which you have passed an an argument is still attached to a running Chrome Browsing Context.
Details
Using Default Chrome Profile for Test Automation will be against all the best practices as the Default Chrome Profile may contain either/all of the following:
browser settings
Extensions
Bookmarks
Apps
Saved Passwords
Browsing History
etc
So the Default Chrome Profile may not be in compliance with you Test Specification and may occasionally raise exceptions while trying to load. Hence you should always use a customized Chrome Profile.
You can find a detailed discussion in How to open a Chrome Profile through --user-data-dir argument of Selenium
If your usecase still warrants to use the Default Chrome Profile you need to ensure that all the google-chrome, chromium or selenium-chromedriver are stopped/killed and you can follow the below mentioned details.
Here you can find a detailed discussion on Selenium : How to stop geckodriver process impacting PC memory, without calling driver.quit()?
Location of Default Chrome Profile
As per the documentation in How to Find Your Chrome Profile Folder on Windows, Mac, and Linux the location for Chrome’s default profile folder differs depending on your platform. The locations are:
Windows 7, 8.1, and 10: C:\Users\<username>\AppData\Local\Google\Chrome\User Data\Default
Mac OS X El Capitan: Users/<username>/Library/Application Support/Google/Chrome/Default
Linux: /home/<username>/.config/google-chrome/default
You need to replace <username> with the name of your user folder. The default profile folder is simply named Default (or default in Linux). However, if you’ve created additional profiles, their folder names are not as obvious. The name you assigned to the profile when you created it displays on a name button on the right side of the title bar on the Chrome window. Unfortunately, the name Chrome uses on the associated profile folder is a generic, numbered name like Profile 3.
If you need to know any of the Chrome Profile's folder name, you simply need to access chrome://version in the address bar and press Enter.
Snapshot:
The Profile Path shows the location of the current profile. For example, the location of my Default profile in my Windows 10 system is C:\Users\Soma Bhattacharjee\AppData\Local\Google\Chrome\User Data\Default. You can select the path and copy it and paste it into File Explorer in Windows, the Finder on OS X or into a file manager like Nautilus in Linux to access that folder.
Sample Code (Windows 10)
Finally, to access the Default Chrome Profile you can use the following Python based solution:
from selenium import webdriver
from selenium.webdriver.chrome.options import Options
options = webdriver.ChromeOptions()
options.add_argument("user-data-dir=C:\\Users\\username\\AppData\\Local\\Google\\Chrome\\User Data\\Default")
driver = webdriver.Chrome(executable_path=r'C:\WebDrivers\chromedriver.exe', chrome_options=options)
driver.get("https://www.google.co.in")
You can find a detailed discussion in How to use Chrome Profile in Selenium Webdriver Python 3

Python Selenium Chromedriver not working with --headless option

I am running chromedriver to try and scrape some data off of a website. Everything works fine without the headless option. However, when I add the option the webdriver takes a very long time to load the url, and when I try to find an element (that is found when run without --headless), I receive an error.
Using print statements and getting the html after the url "loaded", I find that there is no html, it's empty (See in output below).
class Fidelity:
def __init__(self):
self.url = 'https://eresearch.fidelity.com/eresearch/gotoBL/fidelityTopOrders.jhtml'
self.options = Options()
self.options.add_argument("--headless")
self.options.add_argument("--window-size=1500,1000")
self.driver = webdriver.Chrome(executable_path='.\\dependencies\\chromedriver.exe', options = self.options)
print("init")
def initiate_browser(self):
self.driver.get(self.url)
time.sleep(5)
script = self.driver.execute_script("return document.documentElement.outerHTML")
print(script)
print("got url")
def find_orders(self):
wait = WebDriverWait(self.driver, 15)
data= wait.until(ec.visibility_of_element_located((By.CSS_SELECTOR, '[id*="t_trigger_TSLA"]'))) #ERROR ON THIS LINE
This is the entire output:
init
<html><head></head><body></body></html>
url
Traceback (most recent call last):
File "C:\Users\Zachary\Documents\Python\Tesla Stock Info\Scraper.py", line 102, in <module>
orders = scrape.find_tesla_orders()
File "C:\Users\Zachary\Documents\Python\Tesla Stock Info\Scraper.py", line 75, in find_tesla_orders
tesla = self.driver.find_element_by_xpath("//a[#href='https://qr.fidelity.com/embeddedquotes/redirect/research?symbol=TSLA']")
File "C:\Program Files (x86)\Python37-32\lib\site-packages\selenium\webdriver\remote\webdriver.py", line 394, in find_element_by_xpath
return self.find_element(by=By.XPATH, value=xpath)
File "C:\Program Files (x86)\Python37-32\lib\site-packages\selenium\webdriver\remote\webdriver.py", line 978, in find_element
'value': value})['value']
File "C:\Program Files (x86)\Python37-32\lib\site-packages\selenium\webdriver\remote\webdriver.py", line 321, in execute
self.error_handler.check_response(response)
File "C:\Program Files (x86)\Python37-32\lib\site-packages\selenium\webdriver\remote\errorhandler.py", line 242, in check_response
raise exception_class(message, screen, stacktrace)
selenium.common.exceptions.NoSuchElementException: Message: no such element: Unable to locate element: {"method":"xpath","selector":"//a[#href='https://qr.fidelity.com/embeddedquotes/redirect/research?symbol=TSLA']"}
(Session info: headless chrome=74.0.3729.169)
(Driver info: chromedriver=74.0.3729.6 (255758eccf3d244491b8a1317aa76e1ce10d57e9-refs/branch-heads/3729#{#29}),platform=Windows NT 10.0.17763 x86_64)
New error with updated code:
init
<html><head></head><body></body></html>
url
Traceback (most recent call last):
File "C:\Users\Zachary\Documents\Python\Tesla Stock Info\Scraper.py", line 104, in <module>
orders = scrape.find_tesla_orders()
File "C:\Users\Zachary\Documents\Python\Tesla Stock Info\Scraper.py", line 76, in find_tesla_orders
tesla = wait.until(ec.visibility_of_element_located((By.CSS_SELECTOR, '[id*="t_trigger_TSLA"]')))
File "C:\Program Files (x86)\Python37-32\lib\site-packages\selenium\webdriver\support\wait.py", line 80, in until
raise TimeoutException(message, screen, stacktrace)
selenium.common.exceptions.TimeoutException: Message:
I have tried finding the answer to this through google but none of the suggestions work. Is anyone else having this issue with certain websites? Any help appreciated.
Update
This script still does not work unfortunately, the webdriver is not loading the page correctly for some reason while headless, even though everything works perfectly without running this using the headless option.

For anyone in the future who is wondering the fix to this, some websites just don't load correctly with the headless option of chrome. I don't think there is a way to fix this. Just use a different browser (like firefox). Thanks to user8426627 for this.

Have you tried using a User-Agent?
I was experiencing the same error. First what I did was to download the HTML source page for both headless and normal with:
html = driver.page_source
file = open("foo.html","w")
file.write(html)
file.close()
The HTML source code for the headless mode was a short file with this line nearly at the end: The page cannot be displayed. Please contact the administrator for additional information. But the normal mode was the expected HTML.
I solve the issue by adding an User-Agent:
from fake_useragent import UserAgent
user_agent = 'Mozilla/5.0 (Windows NT 6.1) AppleWebKit/537.2 (KHTML, like Gecko) Chrome/22.0.1216.0 Safari/537.2'
chrome_options = webdriver.ChromeOptions()
chrome_options.add_argument(f'user-agent={user_agent}')
driver = webdriver.Chrome(executable_path = f"your_path",chrome_options=chrome_options)

Try setting the window size as well as being headless. Add this:
chromeOptions.add_argument("--window-size=1920,1080")
The default size of the headless browser is tiny. If the code works when headless is not enabled it might be because your object is outside the window.

Add explicit wait. You should also use another locator, the current one match 3 elements. The element has unique id attribute
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as ec
from selenium.webdriver.common.by import By
wait = WebDriverWait(self.driver, timeout)
data = wait.until(ec.visibility_of_element_located((By.CSS_SELECTOR, '[id*="t_trigger_TSLA"]')))

some websites just don't load correctly with the headless option of chrome.
The previous statement is actually wrong. I just got into this problem where Chrome wasn't detecting the elements. When I saw the #LuckyZakary answer I was shocked because someone created a scrapping for the same website with nodeJs and didn't got this error.
#AtulGumar answer helped on Windows but on Ubuntu server it failed. So it wasn't enough. After reading this, all to the bottom, what #AtulGumar missed was to add the –disable-gpu flag.
So it work for me on Windows and Ubuntu server with no GUI with those options:
webOptions = webdriver.ChromeOptions()
webOptions.headless = True
webOptions.add_argument("--window-size=1920,1080")
webOptions.add_argument("–disable-gpu")
driver = webdriver.Chrome(options=webOptions)
I also installed xvfb and other packages as suggested here:
sudo apt-get -y install xorg xvfb gtk2-engines-pixbuf
sudo apt-get -y install dbus-x11 xfonts-base xfonts-100dpi xfonts-75dpi xfonts-cyrillic xfonts-scalable
and executed:
Xvfb -ac :99 -screen 0 1280x1024x16 &
export DISPLAY=:99

strong texttry to add executable path into Service object
options = Options()
options.add_argument('---incognito')
options.add_argument('---disable-extension')
options.add_argument("--no-sandbox")
options.add_argument('-–disable-gpu')
options.add_argument('--headless')
service = Service (executable_path=ChromeDriverManager().install() )
return webdriver.Chrome(service=service , options=options)
its work for me :)

How to start Firefox with with specific profile Selenium Python geckodriver

Here is my code:
profile = webdriver.FirefoxProfile('C:\\Users\\Administrator\\AppData\\Roaming\\Mozilla\\Firefox\\Profiles\\kvycjolb.Prdel')
driver = webdriver.Firefox(profile)
Im not getting any error and firefox starts, but it just does not load with this profile: I have tried changing / to // etc.. but no luck.
This also does not work:
from selenium import webdriver
from selenium.webdriver.firefox.firefox_profile import FirefoxProfile
from selenium.webdriver.firefox.firefox_binary import FirefoxBinary
binary = FirefoxBinary("C:\\Program Files\\Mozilla Firefox\\firefox.exe")
profile = FirefoxProfile("C:\\Users\\Administrator\\AppData\\Roaming\\Mozilla\\Firefox\\Profiles\\kvycjolb.Prdel")
driver = webdriver.Firefox(firefox_profile=profile, firefox_binary=binary, executable_path="C:\\aprog\\geckodriver.exe")
driver.get('https://google.com')
Im getting error:
C:\aprog>testff
Traceback (most recent call last):
File "C:\aprog\testff.py", line 7, in <module>
driver = webdriver.Firefox(firefox_profile=profile, firefox_binary=binary, e
xecutable_path="C:\\aprog\\geckodriver.exe")
File "C:\Python27\lib\site-packages\selenium\webdriver\firefox\webdriver.py",
line 152, in __init__
keep_alive=True)
File "C:\Python27\lib\site-packages\selenium\webdriver\remote\webdriver.py", l
ine 98, in __init__
self.start_session(desired_capabilities, browser_profile)
File "C:\Python27\lib\site-packages\selenium\webdriver\remote\webdriver.py", l
ine 188, in start_session
response = self.execute(Command.NEW_SESSION, parameters)
File "C:\Python27\lib\site-packages\selenium\webdriver\remote\webdriver.py", l
ine 256, in execute
self.error_handler.check_response(response)
File "C:\Python27\lib\site-packages\selenium\webdriver\remote\errorhandler.py"
, line 194, in check_response
raise exception_class(message, screen, stacktrace)
selenium.common.exceptions.WebDriverException: Message: Unable to find a matchin
g set of capabilities

I think the official answer is found in documentation.
Presently that is:
# Custom profile folder to keep the minidump files
profile = tempfile.mkdtemp(".selenium")
print("*** Using profile: {}".format(profile))
# Use the above folder as custom profile
opts = Options()
opts.add_argument("-profile")
opts.add_argument(profile)
opts.binary = "/Applications/Firefox.app/Contents/MacOS/firefox"
driver = webdriver.Firefox(options=opts,
# hard-code the Marionette port so geckodriver can connect
service_args=["--marionette-port", "2828"])

To start Mozilla Firefox with a specific Firefox Profile through Selenium 3.4.3, geckodriver v0.18.0, Mozila Firefox 53.0 and Python 3.6, you need to create a separate Firefox Profile with the Firefox Profile Manager as per the documentation here.
I have created a Firefox Profile by the name debanjan. This profile got stored in this subdirectory:
"C:\Users\AtechM_03\AppData\Roaming\Mozilla\Firefox\Profiles"
The name of the profile (folder) is w8iy627a.debanjan. So while initiating the WebDriver instance we have to pass the absolute path of the Firefox Profile named as w8iy627a.debanjan as follows:
from selenium import webdriver
from selenium.webdriver.firefox.firefox_profile import FirefoxProfile
from selenium.webdriver.firefox.firefox_binary import FirefoxBinary
binary = FirefoxBinary("C:\\Program Files\\Mozilla Firefox\\firefox.exe")
profile = FirefoxProfile("C:\\Users\\AtechM_03\\AppData\\Roaming\\Mozilla\\Firefox\\Profiles\\w8iy627a.debanjan")
driver = webdriver.Firefox(firefox_profile=profile, firefox_binary=binary, executable_path="C:\\Utility\\BrowserDrivers\\geckodriver.exe")
driver.get('https://google.com')
Let me know if this answers your question.

def setFirefoxDriver():
profilePath = r"PathHere"
driverPath = r"pathHere\driver.exe"
options = Options()
options.add_argument("-profile")
options.add_argument(profilePath)
dService = Service(driverPath)
d = webdriver.Firefox(service=dService, options=options)
return d
d = setFirefoxProfile()
d.get('https://www.amazon.com/)
to know profile paths search in your firefox about:support or about:profiles
You can test it loading your own profile and see if cookies are loading, ie: when I go to amazon.com amazon recognizes me.
Notice that you can't be using the same profile in 2 different instances, so if you wanna load your profile to test in selenium you shouldn't be using that firefox profile, but another one.

Always use double backslashes in the path (for Windows paths at least):
profile = webdriver.FirefoxProfile('C:\\Users\\Administrator\\AppData\\Roaming\\Mozilla\\Firefox\\Profiles\\kvycjolb.Prree')
In your code, you use both backslashes and forward slashes.

Python selenium gets 111 socket error

I'm just running the example code of selenium from here:
http://selenium.googlecode.com/svn/trunk/docs/api/py/index.html
The code is :
from selenium import webdriver
from selenium.common.exceptions import NoSuchElementException
from selenium.webdriver.common.keys import Keys
browser = webdriver.Firefox() # Get local session of firefox
browser.get("http://www.yahoo.com") # Load page
try:
browser.find_element_by_xpath("//a[contains(#href,'http://seleniumhq.org')]")
except NoSuchElementException:
assert 0, "can't find seleniumhq"
browser.close()
But It doesn't work for me, here's what it response:
Traceback (most recent call last):
File "test.py", line 4, in <module>
driver = webdriver.Firefox()
File "/usr/local/lib/python2.7/dist-packages/selenium/webdriver/firefox/webdriver.py", line 62, in __init__
desired_capabilities=capabilities)
File "/usr/local/lib/python2.7/dist-packages/selenium/webdriver/remote/webdriver.py", line 72, in __init__
self.start_session(desired_capabilities, browser_profile)
File "/usr/local/lib/python2.7/dist-packages/selenium/webdriver/remote/webdriver.py", line 114, in start_session
'desiredCapabilities': desired_capabilities,
File "/usr/local/lib/python2.7/dist-packages/selenium/webdriver/remote/webdriver.py", line 165, in execute
self.error_handler.check_response(response)
File "/usr/local/lib/python2.7/dist-packages/selenium/webdriver/remote/errorhandler.py", line 136, in check_response
raise exception_class(value)
selenium.common.exceptions.WebDriverException: Message:
...
<div id="content">
<p>The following error was encountered while trying to retrieve the URL: http://127.0.0.1:60106/hub/session</p>
<blockquote id="error">
<p><b>Connection to 127.0.0.1 failed.</b></p>
</blockquote>
<p id="sysmsg">The system returned: <i>(111) Connection refused</i></p>
<p>The remote host or network may be down. Please try the request again.</p>
...

you aren't running the full example. The link you posted contains the following code:
from selenium import webdriver
from selenium.common.exceptions import NoSuchElementException
from selenium.webdriver.common.keys import Keys
import time
browser = webdriver.Firefox() # Get local session of firefox
browser.get("http://www.yahoo.com") # Load page
assert "Yahoo!" in browser.title
elem = browser.find_element_by_name("p") # Find the query box
elem.send_keys("seleniumhq" + Keys.RETURN)
time.sleep(0.2) # Let the page load, will be added to the API
try:
browser.find_element_by_xpath("//a[contains(#href,'http://seleniumhq.org')]")
except NoSuchElementException:
assert 0, "can't find seleniumhq"
browser.close()
this works fine.
The edited version of the code in your question is missing some parts, and therefore fails. Specifically, you are missing these 2 lines:
elem = browser.find_element_by_name("p") # Find the query box
elem.send_keys("seleniumhq" + Keys.RETURN)
That initiates a Yahoo search for "seleniumhq". The results of that search is the content where you want to locate the element.
if you don't do the search, it will fail on:
browser.find_element_by_xpath("//a[contains(#href,'http://seleniumhq.org')]")

When Selenium launches Firefox with
browser = webdriver.Firefox()
the first address it visits is a localhost - 127.0.0.1:xxxxx
If you are using a proxy server, the localhost cannot be visited with the proxy set.
So, first you need to turn off your proxies using:
unset http_proxy
unset ftp_proxy
unset socks_proxy
unset https_proxy
Once you have turned your proxies off, your Firefox should start without any error. But now you need to set your proxies on Firefox. The technique described here works.