1

Recently in my company we had cross browser testing on very big application. We have prepared selenium webdriver script for functional regression testing, however, there is not more scope covered for cross browser testing and that made us to test whole application in all browser (IE 9,10, 11, FF, Chrome, Safari). We have perform this testing as we also include GUI testing like image, button etc., Now, the question is, is there any way we can automate this?

My idea was to get all the element of page and get the location of all elements and save it in xml file, for example let assume that we wrote script for google.com and we got all the list of element with there location (x,y) for IE 9 browser, now run same script with rest all the browser and save the element location and than compare each element location with different browsers out, this how we can get the details information about element location in different browser, we can also use size, isdisplay, isenabled method to verify element. Using location method we can find which element does not exist or not located on correct place...

This are my thought, can I get any input from you?

Thanks!

Ajinkya
  • 22,324
  • 33
  • 110
  • 161
Karim Narsindani
  • 434
  • 3
  • 14
  • 38

2 Answers2

3

Your thoughts are mostly correct, except the fact that you cannot compare exact locations of elements in browsers (they will usually differ) but you can compare how the elements are related to each other (e.g. element x should be on the right side of element y, etc.).

I know of two commercial solutions that offer functionality similar to what you described: Mogotest and CrossBrowserTesting. They both offer free trials, so have a look at them. Browsera claims to have such functionality too, but I've never used it so I cannot comment on it.

If you prefer to implement such system by your own then there are two sources you may find helpful:

JacekM
  • 4,041
  • 1
  • 27
  • 34
  • Hey JacekM, Thanks for your input, and yes I would like to implement this and to create dynamic system. Hoping to jump all the hurdles while implementation. Well thanks for you response and links!!! – Karim Narsindani Mar 24 '14 at 07:20
0

There is quite high level of false positive issues with this approach. There are multiple cases where you'd see a given button reported to be 3 times larger on some browser while it's not. Apart from this it's a good one since it would also tell you the difference

Slavy Slavov
  • 11
  • 2