Selenium: Bug - multiple calls to driver.get(url + '/#!/home'); fail

[ORESoftware]

I posted this issue on SO:

https://stackoverflow.com/questions/44208949/selenium-routine-works-the-first-time-but-subsequent-runs-fail/44209526#44209526

It looks like if I call this more than once

driver.get(url + '/#!/home');

it will not work after the first time, as described in the SO issue. This appears to be a bug. As a workaround, I have to explicitly call refresh().

Is there a specific issue tracker for the NPM/Node.js Selenium webdriver?

[akshat62]

It should work. I have tried it. Provide more code.

Sometimes it happens that concatenation doesn't happens successfully.
Check it once. Split your code.

Use
String url= cdthost + "";
System.out.println(url);
driver.get(url);

[p0deje]

Please provide concise reproducible test case so that we could act on this issue.

See CONTRIBUTING.md

[thomasbalsloev]

I see the same issue I think, but only with Firefox.
Navigation to the same URL as the current URL makes the test stall.

Our set up:
Ruby bindings with Selenium driver 3.4 standalone JAR.
Gecko driver 0.17.0 (64 bit)
Firefox 54 (32 bit)
Windows 7 Ultimate (64 bit)

We have the same Selenium ruby code run with three browsers. Chrome and IE run with no problems.
I'm not certain this stems from the Selenium driver at all.
It might be the gecko driver and/or FireFox itself.

I used the workaround that @ORESoftware has written on the SO issue.

Hope this helps someone...

Thomas.

[ORESoftware]

@P9GIT yep, I was sure of it, it's a real problem, thanks for verifying it. I was using Firefox when this issue occurred. Can I get an upvote on the SO solution that worked for you? ;)

[ORESoftware]

@p0deje you have all the information you need to fix this issue? should be clear

[p0deje]

you have all the information you need to fix this issue? should be clear

Not, I don't. I don't see this issue locally when trying to navigate to the same URL. Please give us a script that can be run locally to reproduce the bug.

[thomasbalsloev]

@ORESoftware

Sure, upvoted.
Though I have zero reputation points being a newbie so their system won't show the upvote (it's recorded though).

[thomasbalsloev]

@p0deje @ORESoftware

I did some more tests and got these results. When visiting already loaded page:

DOES NOT WORK:
http://www.ford.com/#here
http://www.ford.com#!/here
http://www.ford.com#here/more

WORKS:
http://www.ford.com
http://www.ford.com/here
http://www.ford.com/here#here

pseudo-code:

driver_instance.navigate.to("http://www.ford.com#here/more")
driver_instance.find_element(:link_text,"Vehicles")
driver_instance.navigate.to("http://www.ford.com#here/more") <--- stalls here when the "bug" happens

It seems the fragment composition is the key here.

We see our failure on this URL path+fragment: /portal#issues/snoozed
Tested this on several other of our fragments of this form and they all fail in the same way.

Thomas.

[p0deje]

Thank you, I can now reproduce the bug, which seems to be GeckoDriver issue.

I've created mozilla/geckodriver#790 which you can track from now. Let's hear back from Mozilla guys.