SOLVED

UI Automation (edge chromium)

Copper Contributor

What is the expected method to UI (ux) automate the new Edge (insider) browser? 

10 Replies
best response confirmed by Eric_Lawrence (Microsoft)
Solution

@BobbyCannon we plan on having an MSEdgeDriver very soon, and you can also use puppeteer and headless. In addition, you can use ChromeDriver for the time being by passing in the path to msedge.exe in a ChromeOption. Is there anything specifically you would like? 

@johnjansen I have an automation open source project that I want to add support for Edge. 

 

I would suggest, if you can do this, to stick with the Chrome way of automating. It's uber nice to not have to download yet another thing to get automation working. I never supported Edge due to the fact that I'd have to instruct TestR to download the "driver" to enable support. Just too much friction. 

 

One of the benefits of TestR over other automation frameworks is that it just works with very little friction. IE and Chrome just works and FireFox has very little configuration requirements. You can reference one nuget package, add a few lines of code, and boom you are automating browsers and desktop applications.

@BobbyCannon whatever you are doing for Chrome should work for the new Edge, so that should be fine. How are you automating Chrome?

@johnjansen any plans to support IE like automation? we use winforms to do browser automation that launches IE and navigates through website for our customers. Right now, using a WebDriver seems to be a solution more for test automation more than end user automation.

 

Additionally, when I had tried Selenium with Old Edge, I had to match it to the Windows version which is not in our control. Second, if I remember correctly, I could not release the hook to the selenium launched browser window as ending my program also closed the browser.

@CodeDJ TestR would be a great solution for testing your WinForms + IE project. I could help you get started if you would like?

 

TestR is Open Source (MIT) and you can use the same API to automate your desktop (winforms) application and whatever is running in the browser. 

@BobbyCannon thanks and just to be sure. Are you suggesting to use TestR for browser automation from WinForms on end user machines? If not, sorry for the confusion. The requirement is, my app should launch a new browser, navigate through a website. The user can then take over and decide what to do from there on, including whether to close the browser or navigate away to a totally different website.

 

Today, we do this using IE. Assuming IE is going to be replaced by new Edge, we will have to find an alternate way of doing this.

 

As mentioned in the previous post, I tried Selenium with Chrome/old Edge but could not release the browser once done navigating i.e. after our automated navigation, the browser and my app should be totally independent.

TestR can attach / detach from existing browsers. Interested in how your app is starting the browser and automating it.

The intesting thing is you could implement TestR in your application and do that automation in your application which would immediately get you support for IE, FF, and Chrome.

Then you could test everything external also using TestR in your "unit test" project. We should schedule a time, I don't mind showing you a demo via Skype or Hangouts.

@BobbyCannon thanks for the info. I will try TestR to see if it meets our needs.

 

We use SHDocVw.InternetExplorer currently to do the automation.

@CodeDJ I did want to add that we have heard the request about ending your testing application also closed the browser. We have been looking into that. I will say that it sounds like your winforms app for executing the browser is something that WebDriver is meant to do. The goal of webdriver is to replicate the actions of an actual human as we best we can.

Hi , Could you please help me , i need exactly same , i want to automate lot of stuff using edge browser like controlling windows applications
1 best response

Accepted Solutions
best response confirmed by Eric_Lawrence (Microsoft)
Solution

@BobbyCannon we plan on having an MSEdgeDriver very soon, and you can also use puppeteer and headless. In addition, you can use ChromeDriver for the time being by passing in the path to msedge.exe in a ChromeOption. Is there anything specifically you would like? 

View solution in original post