The script is crashing on Chrome version 103 only it's still running on 102. In one instance, it ran but crashed while executing a loop for its fifth time. It is selenium + python. What should I do?
this issue came with chrome 103. Remove the user-data-dir option for solution.
It was fixed for me when I removed the user-data-dir option.
It looks like this problem will go away when chrome 104 is released, we'll be dealing with workarounds until chrome 104 comes out.
https://bugs.chromium.org/p/chromedriver/issues/detail?id=4121
https://github.com/SeleniumHQ/selenium/issues/10799
New Answer
Chrome Version 103.0.5060.134, selenium is now working as normal
Old answer
I actually:
Uninstalled Google Chrome
Restarted the computer
Installed an older version of Chrome(whilst the Internet is OFF)
Run MS config & disable Chrome auto-update
Restart the computer.
Now it runs just FINE.
NB: This can make your computer vulnerable to attacks so don't use your main PC. Use Isolated Environments
.
Related
By some project reasons I have to use Selenium 4 (4.0.0-rc3) and Chrome version 71.
Of course the current chromedriver (version 94) doesn't work with old chrome.
I have downloaded the source code of chromium project for debugging and try to understand how actually driver works with selenium. Probably I will try to make my own driver version but at this moment it seems like an inefficient solution.
I am really stuck with this and will be happy to get any advice.
Is it possible just to configure driver by capabilities (or any parameters) for working with old chrome?
The Chrome version you're targeting Chrome 71 is already deprecated. Using Selenium 4 with this much obsolete version of Chrome will not work and I think apart from creating your own code piece, there is nothing much you can do.
Using such an obsolete piece of software, in this age of internet, will open a pot of bugs for you, and if you use this to connect to internet, then chances are people may exploit any hidden security vulnerabilities to attack your application.
I would strongly suggest you to talk within your project team to use an updated version of Chrome.
ChromeDriver 2.46 will work with Chrome 71.
I'm debugging a React Native 0.61.5 app with Chrome and usually when code stop running on a breakpoint and I hover a property, I can see all the details of it in a popup window. But for two days now I can't get back this functionality. I restarted computer, cleaned Chrome and Metro bundler, tried with virtual or real device but it's still not working. Did I deactivated something accidentally ?
Update:
The issue seems to be resolved with the new Chrome update 79.0.3945.117.
Yes, with Chrome 79 this functionality was broken. This is the thread where people are reporting this bug:
https://support.google.com/chrome/thread/22882968?hl=en
(and also in the link from the comment to this question)
And this is a thread where Chrome developers are tracking this issue and trying to fix it:
https://bugs.chromium.org/p/chromium/issues/detail?id=1033022
In Chrome Canary (81) this feature is working, and it's reportedly working in the Chrome Dev version (80). Other alternatives are Firefox (also people say that it's working in Opera and other browsers) and debugging via VSCode.
It seems they aren't going to fix it until the version 80, which isn't expected to be released until February (from what people say in that thread: https://support.google.com/chrome/thread/22882968?hl=en).
Also there was this change: https://chromium-review.googlesource.com/c/chromium/src/%2B/1835357
which made variables appear on hover only inside the innermost scope. Update: this feature (to evaluate outside the innermost scope) will probably be restored:
(from https://bugs.chromium.org/p/chromium/issues/detail?id=1033022)
"This undoes
http://chromium-review.googlesource.com/c/chromium/src/+/1835357,
which limited the eager popover evaluation to the innermost scope, which
is the only scope for which we can produce reliable results right now.
For the outer scopes, it might still produce the right results, but we
can't guarantee that. But the feature is still very useful and so we are
restoring the behavior here, and will follow up with a proper fix and UX
later".
Even with last Google Chrome update I was not able to resolve this issue. The issue comes from the Chromium version of Google Chrome.
You need the version 81 of Chromium who fixes the bug.
I downloaded the last Chromium browser with Chromium 81 and it's working for me https://chromium.woolyss.com/download/fr/
I have IE 11 installed, but I need to run a few tests in IE 8, 9, & 10 as well. How can I force the compatibility mode when running a test.
FYI, I did try F12, and set the mode to 8. It works manually, but not through selenium.
Thanks!
I ran into exact same issue recently. Unfortunately, there is no way unless you change page source. See this. However, I can tell you what I have done. I used VMS of course. Luckily, microsoft also provides trial versions of VMS with different versions of IE. visit this. After getting the VMs ready I installed NUNit Console. I then simply copied and loaded the test dll on that and just run.
I have a client who had a script working and now stopped working. I run it in my computer and it works (i runned it on Linux and Windows). My client runned it on Mac and Windows.
I am new to Selenium, but the part of the code that crashes is where Selenium takes place, it gets the Firefox to open but not load the page. I think it migth be a problem with the webDriver of Selenium that must be updated.
I told him to install pip in his Mac and uninstall Selenium and reinstall it, is this how it is done in OS machines? Is there any issue related to my problem?
Thanks
It would be helpful if you had specify Selenium version and also which browser are you using and it's version too. But the behavior you described seems to be caused by having newer version of browser which the Selenium version you using doesn't support. So upgrading to newest version of Selenium (2.35) will likely make this problem disappear.
I am using selenium-server , selenium rc for UI testing in my application . My dev box is Windows with FireFox 3.5 and every thing is running fine and cool. But when i try to run selenium tests on my build server which is a linux box , i get this error
Caution: '/usr/bin/firefox': file is a script file, not a real executable. The browser environment is no longer fully under RC control
basically the selenium-rc is not able to find the firefox executable on linux(actually its a .sh file here), which is present in the case of Windows .
Has anyone faced this error before. Please let me know for any pointers.
Thanks for any help
~Pratik
Thanks!
Pratik
This is occurring because Selenium wants to start the FireFox binary directly, rather than delegate to the normal launch script. Before checking for "firefox" (which is a script on many installations), it checks for "firefox-bin". You can circumvent this problem by creating a link named "firefox-bin" that points directly at your desired version of the firefox binary.
This is described under "Ubuntu fix" (though it actually applies to other Linux'es as well) in the Cucumber Wiki at:
https://github.com/cucumber/cucumber/wiki/Setting-up-Selenium/612fae3f39528be39594d145ec12a4747855e030
Hey, I have Selenium running on Karmic with no issues. I can't quite remember how I did it, but I think it was a case of grabbing a more recent version ot the selenium-rc-serve.jar.
I also remember looking at this post: http://tero.tilus.net/rutinat/2009/08/18/cucumber-running-selenium-on-debian-lenny/
Hope something there points you in the right direction!