Help required with understanding Cross-browser Testing

Question

Based on my knowledge so far of cross-browser testing may I ask the following:

1. Is it true that at majority of the companies it is not performed, since most of the time when something working on one browser and not working on another, it is because something is not configured or updated properly in that browser?
2. Is it true that even when it is worth doing C.B.T., any time spent on it is time lost on functional testing and most of the defects have to do with the code and little to do with a browser?
3. When I was doing manual testing, it was required of me to test in two browsers at only one of the places I've worked at. Would it be safe to say that in automation testing C.B.T. is as common (or rather as uncommon) as it is in manual testing?
4. Most of my experience in manual testing has been through I.E. Is it true that if an application is tested in only one browser using Selenium, it is standard to test it in Firefox?

Answer 1

"Is it true that at the majority of companies it is not performed"

Well I haven't done a survey, have you? As far as I know web development companies usually do Cross-browser Testing (C.B.T.). They have to make sure the websites and web applications they develop look and work in the same way in various browsers. This is to make sure that all users have a good experience on the site or application and want to visit again. Plus, they have to do this is because they don't control which browser and which version of that browser their end users will be using. The most they may do if an end user reports an issue is ask them to upgrade to the latest version. Mostly the companies test of few widely used browsers like FireFox, Chrome, Safari, Opera and I.E.

"since most of the time when something working on one browser and not working on another, it is because something is not configured or updated properly in one browser"

Unless an until your end user has some technical knowledge and requires specific configurations and plugins set up in their browsers, they will install the browser with default settings and use it that way only. So, no its not about "something is not configured or updated properly in one browser". Its about something wrong in you code.

"Is it true that even when it is worth doing c.b.t. any time spent on it is time lost on functional testing and most of the defects have to do with the code and little to do with a browser?"

No. Time spent on C.B.T. is not time lost on functional testing. If it is then the problem is with your analysis and estimation. You should have estimated C.B.T. and functional testing separately. "most of the defects have to do with the code and little to do with a browser" When you say C.B.T. are you testing your website or web application's compatibility with various browsers or testing the browser? Yes most defects have to do with problem with the code and very few occur due to problem with the browser!

"Would it be save to say that in automation testing c.b.t. is as common (or rather as uncommon) as in manual testing?"

This totally depends on the individual thinking. If you are comfortable with writing automation script to test a website or web application of various browsers then it will be common for you.

"Is it true that if an application is tested in only one browser in Selenium, it is standard to test it in Firefox?"

Selenium provide built-in support for FireFox. That is why people prefer FireFox. You can install and configure drivers for other browsers separately and then use them. But, with C.B.T. people usually test the design compatibility more than the functionality. Selenium will help you test the functionality of your website or web application and not the design.

Answer 2

Answers:

1. No - I have seen majority of the companies do cross-browsing testing. One of the cause could be browser configuration and updation but there are lot others too involved in that.

2. Yes - I do think one should prefer functional testing to complete and then can move towards cross-browser testing. But surely not at the start.

3. Yes - Cross-browser testing included in both and should be carried out in both automation and manual testing.

4. No - Its not mandatory to go for firefox. But it has good support available. Selenium execution observed good with and on firefox. We can use other browsers like chrome or IE with selenium

Its open if you would like to add points!

Answer 3

1) It is very common for websites to be tested upon multiple brands and versions of browsers.

2) Where we work, we conduct testing the primary browser used by our customers, we then schedule browser testing on anything our customer use with a cut off of 5%. This initial created a problem for the team as the amount of resource need to test a change increase. To combat this we separated our testing out in to two areas. Functional and Usability and rendering. Functional Testing is carried out by our systems testers and Usability by our digital content testers. We then decided that even this approach did not solve the problem and started automating our tests using saucelabs. Now we test on many variations of browsers, but only spend any hour of virtual time.

3) Exceptionally common as is highly recommended. I would have a look on youtube for saucelabs.com and appium (if you are testing mobile apps) ever since we started this process, we now have usable websites that function and we are not loosing resource.

4) No, If you have to do the bare minimum. IE (One Version behind normally), Firefox and Chrome. I would also recommend Safari on IOS, but it depends on what your users are using as it can vary. Check your Google Analytics stats and start to create a profile/policy of what your customers use and browsers and devices they may use in future. This than allows you to know what to test and what risks you are taking.

Answer 4

Even places which use automated regression test, do a lot of testing manually. Developing (and maintaining) test is a significant time commitment, is not free. So it makes sense to automate most often used workflows, but test manually other functionality. Of course it depends of how complex you web app is.

Priority of browsers in CBT (or anything else) should be in order of preference of YOUR users, not guided by opinion from some random person from internet. If 90% of your users use Safari on iPad, how would be testing in FF and Chrome be relevant? Measure relevant metrics and make your own decisions.

Answer 5

Most of the health product companies do testing in all browsers irrespective of the Operating System