After Selenide 6.0.1 version they deleted Configuration.startMaximized. How can we maximize browser now?
Just use the pre-defined size of the browser, it helps you to avoid unexpected behavior in tests, especially on remote runs.
Configuration.browserSize = "1920x1080";
Related
I appear to be having the exact opposite problem many other people have - in that my Selenium tests in Ruby will close the browser window at test end, no matter what the end result is. Pass or fail, it will always close the browser. I would like to stop this.
Context:
Previously I coded tests in Java using IntelliJ IDEA. Browser windows for Selenium tests in this case would NOT close at all period unless you used driver.quit(). This is actually quite useful as it means that the browser window would stay open if the test failed - which meant I could look at where it stopped in the browser and help figure out why it failed. This was also useful for test writing as it meant that I could essentially pick up where I left off to write the next block instead of having to keep a parallel tab going in another browser by hand to get the next set of selectors in the given screen.
I've found in Ruby using RubyMine that the browser will close when the test ends in any capacity. This is a bit of a problem. While technically I could take a screencap on failure, it'd mean that I'd have a harder time retracing why it failed (back button on browser, typing in fields to work out if a quirk in our UI caused it, etc). And of course, screencaps take up hard drive space. ;)
I've tried the detach=true (and True) command switch for Chrome and that has not worked.
Setup:
IDE: RubyMine
Gems: selenium-webdriver
Browser: Chrome, using ChromeDriver. (In Ruby this just involves using WebDriver)
OS: OSX
Not headless, using no other frameworks/testing environments. It is quite literally a few lines of setup and then hitting the run button in RubyMine in a bog-standard .rb.
Summary:
I haven't been able to find any existing questions here or in other places online for Ruby specifically for keeping a Selenium test in RubyMine from closing the window on test end. "Test end" in this case refers to success (reaching the end of the .rb) or failure (Tracebacks, in this case). I would prefer that the window would stay open until it'd hit a driver.quit line. Is there any way I can set this up?
Thank you very much. I hope this isn't redundant. I also hope this will help other testers in the future :)
You can use the :desired_capabilities to set this flag:
caps = Selenium::WebDriver::Remote::Capabilities.chrome("goog:chromeOptions" => {detach: true})
driver = Selenium::WebDriver.for :chrome, desired_capabilities: caps
Note that old examples will be using "chromeOptions", but for newer Chrome versions, it will need to be "goog:chromeOptions".
#JustinKo has posted an answer here
Chrome detach option is no longer working
You can pass Chrome Options using the following format:
browser = Watir::Browser.new(
:chrome,
'goog:chromeOptions' => {detach: true}
)
The answer from #Justin Ko worked for me. However at the time of posting I am getting a warning telling me to use
capabilities: caps
instead of
desired_capabilities: caps
as it is now currently deprecated at the time of posting.
I do not have enough reputation to be able to comment so I am posting this as an answer.
I have written a code for Web page. The process requires me to click on a weblink which opens up a new window, then perform some operations on the browser window. Then I close the new browser. This is repeated multiple times in the code. All the elements on all the browser windows are normally identifiable using the object spy. However, intermittently during run time when a new browser window opens up the elements on the page are not getting recognized (hence it throws errors). When i go into the debug mode and try using the object spy the maximum identification i can capture is Browser(<>).Page(<>). Nothing in the page is getting recognized.
Now if i close this browser and reopen it and check again, the elements on the page are getting captured by the object spy and i can continue with my script execution. Sometimes I have to close and reopen multiple times for it to work.
Is there any way to handle this scenario. check for object identifications on the run time maybe. Dunno if it this is any relevant but i am not making use of the OR in my project.
Thanks in advance.
This sounds like a bug in UFT and you should contact HP's support.
A workaround if you know where the problem is probable to appear is to add Browser("<name>").RefreshWebSupport. This is an undocumented feature of UFT that sometimes helps in cases like this.
When I run selenium tests that use Chrome as the browser, the tests hang. The reason is that, since the browser is running as the SYSTEM user, it continually pops up Chrome's prompt for search engine choice. If I run the selenium server interactively, and as a logged-in user select a search engine, it will enable the tests to run. However, the next time I run the tests I get the prompt.
Is it possible to avoid this behavior?
This happened with me on Firefox as well. What i did was run Firefox as administrator and answered all the prompts that one time.
Next time when I ran the selenium scripts, those modals did not show.
Not sure if this would directly relate to Chrome, but definitely worth a try.
Let me know if it worked?
I can suggest you 2 options:
a. Open Chrome with a specific profile.
b. Write and run setup routine before any tests, which will open browser settings page and do whatever you need too:
from selenium.webdriver import Chrome
SETTINGS_PAGE_URL = 'chrome://settings/browser'
SEARCH_ENGINE_DROPDOWN_ID = 'defaultSearchEngine'
SEARCH_ENGINE_CHOICE_XPATH = '//option[text()="Google"]'
browser = Chrome()
browser.get(SETTINGS_PAGE_URL)
dropdown = browser.find_element_by_id(SEARCH_ENGINE_DROPDOWN_ID)
option = dropdown.find_element_by_xpath(SEARCH_ENGINE_CHOICE_XPATH)
option.click()
browser.get('http://wherever.you/need/to/go/next/')
I'd use option a.
I'm using Selenium Client 2.4.0 on Mac 10.6.6 with Firefox 5. Using the WebBackedSeleniumDriver, I'm running a "selenium.getEval" command that causes a Firefox warning,
"Warning: Unresponsive script.
A script on this page may be busy, or it may have stopped responding. You can stop the script now, or you can continue to see if the script will complete.
Script: resource://fxdriver/modules/utils.js:9161"
The value of "dom.max_script_run_time" about:config was "0", which should disable the above dialog altogether. Yet, I still get the dialog. Is there any way to prevent the warning dialog from appearing?
dom.max_script_run_time is the right preference but it only applies to web pages. Browser UI (and extensions like fxdriver are part of it) are restricted by the preference dom.max_chrome_script_run_time however (default value is 20 seconds). You should set it to 0 as well.
Documentation: https://developer.mozilla.org/en/Preferences/Mozilla_preferences_for_uber-geeks#DOM_preferences
Regardless of what you have setup in your profile, during startup, Selenium sets both values to 2417483647 to try and get around the browser warning. However, because the value is so large, FF ends up ignoring it and using the default value of 10/20 instead. This is true even if you're pointing Selenium to use your profile as the template.
The best way I've found to get around this is to specify
-timeout nnnn
to the Selenium Server startup args. This sets both the server and client (browser) timeout values.
Although this thread is quite old, the problem still exists with current selenium and firefox builds. I've got these really annoying messages quite a long time now, without a good fix. Asking the devolopers / mailing list / google usually results in the following answer:
The javascript used in your application is too fat or buggy, improving your scripts will help.
As this is no option in a larger company, when you depend on a framework you have no access to, i decided to search for the root cause for myself.
The core of the problem is the fact, that selenium overrides profile settings if you specify the -timeout nnnn parameter. So creating a custom firefox profile template and setting the dom.max_script_run_time and dom.max_chrome_script_run_time will not work here.
As soon as you specify the -timeout parameter, these two settings are overriden with the value you provide to the parameter. After hours of debugging and testing i noticed some facts:
If you don't specify -timeout, firefox runs for exact 30 minutes, without one script timeout. After that, firefox gets killed by selenium with a SeleniumCommandTimedOutException
As soon as you specify -timeout (no matter which value), the script timeout appears after several seconds or minutes. These messages are independent to the timeout-value.
The -browserTimeout parameter isn't usefull as i haven't found where this parameter is used in the source.
As we have some testsuites that run longer than 30 minutes we have 2 options to fix this behaviour:
Rewriting our testsuites and splitting them to run within the 30 minutes window
Patching selenium to run longer than 30 minutes
Do not use the -timeout parameter.
So choose for yourself which option is better. I created a small and simple patch for the HTMLLauncher.java to allow 90 minutes instead of the default 30.
diff --git a/java/server/src/org/openqa/selenium/server/htmlrunner/HTMLLauncher.java b/java/server/src/org/openqa/selenium/server/htmlrunner/HTMLLauncher.java
index c2296a5..310b39f 100644
--- a/java/server/src/org/openqa/selenium/server/htmlrunner/HTMLLauncher.java
+++ b/java/server/src/org/openqa/selenium/server/htmlrunner/HTMLLauncher.java
## -146,6 +146,16 ##
launcher.launchHTMLSuite(suiteURL, browserURL);
sleepTight(timeoutInMs);
+ // SFR, Patch 2013-10-17: To get rid of the damn SeleniumCommandTimedOutException
+ // we allow the Suite to run 3 times as long as per default (30 min -> 90 min).
+ if(results == null) {
+ log.warning("SFR, Patch 2013-10-17");
+ sleepTight(timeoutInMs);
+ }
+ if(results == null) {
+ log.warning("SFR, Patch 2013-10-17");
+ sleepTight(timeoutInMs);
+ }
launcher.close();
I'll upload a pre-compiled jar with the above patch if necessary.
Go the hidden configuration page in Firefox by typing about:config in the address bar . (make sure that you are doing this for the profile you are using for selenium) In the 'Filter' box, type script_run_time.
This will narrow the options to dom.max_script_run_time and dom.max_chrome_script_run_time. Right-click it and choose Modify. A box pops up. Change the number to something bigger like 40. This is the maximum time a script can run before Firefox considers it 'unresponsive'. If you can’t find the string in the about:config page, create it by right-clicking anywhere and then choose New—> Integer and enter there name and values (when asked)
Set it to a very large number instead of "20"?
The dom.max_script_run_time setting is in seconds. Changing it from 10 to 20 just doubles the amount of time to wait. A value of 0 will disable it, but this could result in a run-away script locking up your browser. You might just use a really large value.
More details here:
http://kb.mozillazine.org/Dom.max_script_run_time
This issue appears for me when I do Selenium calls while web application is running some own crappy slow JS. Even if I catch the popup and retry after few seconds, FF is not responsive to Selenium anymore. The solution is to just put sleep 10 before otherwise any Selenium call would result in popup.
I am trying to get selenium tests to run. Yet every time I try to run a tests that should run IE I get a error on line 863 of htmlutils.js It says that I should disable my popup blocker. The thing is I went to IE tools-> turn of popup block.
So it is disabled and I get this error.
Is there something else I need to disable. I actually don't even know what version of Internet explorer it is running since I am using Windows 7 Pro 64bit version. So when I do use IE I use 64bit version but I am under the understanding if the site or something like that does not support 64bit it goes to 32bit.
So not sure what I need to do it to make it work.
This is the lines where it does
function openSeparateApplicationWindow(url, suppressMozillaWarning) {
// resize the Selenium window itself
window.resizeTo(1200, 500);
window.moveTo(window.screenX, 0);
var appWindow = window.open(url + '?start=true', 'selenium_main_app_window');
if (appWindow == null) {
var errorMessage = "Couldn't open app window; is the pop-up blocker enabled?"
LOG.error(errorMessage);
throw new Error("Couldn't open app window; is the pop-up blocker enabled?");
}
Where is this log.error message stored? Maybe I can post that too.
I had a similar problem on Vista and IE8
I would get the same error message
Couldn't open app window; is the pop-up blocker enabled?"
Running my remote control as Admin wasn't an option for me, and also a poor idea from a security perspective.
So in the end I manage to solved this by changeing browser from "*ietha" to "*iexploreproxy"
grid_configuration.yml
hub:
port: 4444
...
- name: "Internet Explorer 8 on Vista"
browser: "*iexploreproxy"
...
Alternatively, you can change browser string from the code:
ISelenium selenium = new DefaultSelenium("localhost", 4444, "*iexploreproxy", "http://www.google.com/");
Works like a charm.
The only question remaing is if this somehow affects the outcome of the test cases. So far no, but I'll update this answer in case that would happen.
I ran into this on Windows 7 64bit.
My solution was:
Disable popup block. - Select "Tools/Popup Blocker/Turn off pop-up blocker"
Disable IE protected mode. - Untick "Tools/Internet Options/Security/Enable protected mode"
It'd be better just to disable protected modes for known trusted hosts/addresses. I'll leave that as an exercise for the reader.
I was experiencing the same problem. I ran the Selenium RC server as an administrator and everything worked fine.
I, too, am experiencing this very problem on a Windows 7 64bit box, trying to run Selenium on it to test and ASP .Net MVC application, written in C#.
I am still trying to work out the answer for myself, but I thought I'd post here to tell you of a little progress I have made in getting something to work, albeit in Firefox instead of IE.
Here's the line I changed:
selenium = new DefaultSelenium("localhost", 4444, "*chrome C:/Program Files (x86)/Mozilla Firefox/firefox.exe", "http://www.bbc.co.uk/");
I would ideally like for this to work in Internet Explorer 8, but if for the moment, I can begin getting tests working and later change over to use IE again, then great.
Hope this helps for your problem with it all.
I had the same problem on Windows 7 64bit IE8. The first step was to disable the IE popup blocker. Then, I got a message in the status bar saying that "Pop-ups were blocked on this page. Press the 'Ctrl' key to allow the pop-ups".
It turns out that the Google Toolbar was providing this feature. Disabling it solved the problem. View > Toolbars > Google to toggle.
John.
If you happen to be doing this from JavaScriptMVC, there is a reference you need to change in \jmvc\plugins\test\drivers\selenium.js:
1) Change iexplore to iexploreproxy and you should get better results:
msie : (/iexploreproxy/i).test(browserStartCommand),
2) At this point, you'll find that you still get the popup error, but a separate instance of IE has started. Leave that IE window open and restart the tests, but not Selenium.
3) Next, the windows should show up in the right place, but IE gives the annoying block active content warning. Allow the content to run and restart the tests, but not Selenium itself.
This is super clunky, but it at least gets you past that part. If I find more methodical ways to do these things I'll update as needed.
I have had the same problem and have found another solution which works for me. Just use the *iexploreproxy setting in the browserString.
I used:
selenium = new DefaultSelenium("localhost", 4444, "*iexploreproxy C:/Program Files/Internet Explorer/iexplorer.exe", "http://www.bbc.co.uk/");
I hope that works for others too :)
You can start the test when you disable the Security mode of Internet. Don´t know the correct name for it, but in dutch it is beveiligde modus.
I tried modifiing the security settings to dublicate this security mode, but couldn´t find the correct setting for it. It must therefor block more then you can set manually.