Ask IH: How you guys perform cross browser testing of your websites or web apps?
4
votes
I am keen to learn how you guys perform cross-browser testing these days within your budget. Have you any recommendations?
1
For desktop browsers, I test each using the actual browsers. For mobile, however, chrome has a set of emulators that can be found via their inspect tool. Honestly have saved my life when I had to make changes in production.
1
Is current browser version is sufficient for ensuring cross browser compatibility of your website?
People can have different type of browsers. browser versions, Operating systems and resolutions. If we calculate it will be around 2000 combinations of configuration. So I am asking to ensure cross browser compatibility across 2000 combinations.
2
I wouldn't even try to accommodate every browser / browser version combination out there. Your brain will simply explode!
Older browsers may not even support certain features that form an important part of your application. Even browser manufacturers don't support all versions of their software. At a certain point, they end of life them.
The client has a responsibility to upgrade his or her browser at some point, and that is as true of an individual consumer as it is of a business client. No-one can expect everyone to support their technology stack from now until the end of time. You know that is true ... otherwise we would all still be using DOS and the browser question simply would not arise.
There's another question to consider. How many people are actually using the different versions of each browser? Are you really going to put as much development effort into supporting a browser that is used by 4 people who may not even become customers as you are into supporting a browser used by 1,000 people who most definitely are?
This link might help you to gain a little perspective:
As you can see, for each browser, the version(s) being most used are highlighted. Older versions are now barely used at all.
If I were you, I would develop according to the latest broadly supported standards, make sure your app works properly across the new browsers and simply tell other people "tested on x, y and z. If your browser is older than this, please upgrade at [provide links]".
Otherwise, the one thing I can promise you is that you will never launch a product and the practical difficulties of trying to support 2,000 or more combinations, most of which no-one will ever use, will just overwhelm you and cause you to throw in the towel.
You do not need the sort of people who stubbornly refuse to keep up with the times as your customers. They will cost you far more than you will ever earn from them.
For the ones who are a little behind, chances are your app will work perfectly well. It is only the big changes in browser technology which ought to concern you. And there is nothing wrong with selling the idea of an uphgrade (they're free!) to customers who are slightly behind. You sell the idea on the fact that they can use features of your app (real or imagined) that their current browsers can't process. It is very common on websites to see links to the browser download pages of the different manufacturers. A few seconds' work on your clients' part will save you years of development frustration.
Don't be scared to tell your customers what you need from them. Most will be happy to oblige once they understand what is being asked.
1
Thanks for your reply@thomasm1964
This article will help understand why we need Cross-browser testing.
I have to agree with @thomasm1964, Chasing after absolute perfection often leaves us at a stand still. Focus on a realistic set of parameters in a specific age-range and ship 😀
1
True.
I'm not denying that you need cross-browser compatibility. What I am saying is that you do not need to be compatible with every browser that ever existed. That is simply an impossible thing to do and will only become more impossible as browsers, client-side languages and CSS continue to evolve and screen resolutions / sizes change.
What you have to do is decide upon a sensible set of compatibilities and work to that set.
Anything which is not being used by a significant number of potential users should not be supported. It is as simple and as brutal as that.
If people are not using specific browser / screen combinations, how much business are you going to lose by not supporting that combination - and how much will it cost you to continue to support it?
Don't be frightened to set out your own boundaries in accordance with your own real requirements.
1
Hey! Are you talking about exploratory tests (ensuring that new features work across browsers and generally clicking around), or regression tests (ensuring you didn’t break old features with each change)?
1
Yes you got it right. All the functional testing which includes exploratory and regression.
For desktop browsers, I test each using the actual browsers. For mobile, however, chrome has a set of emulators that can be found via their inspect tool. Honestly have saved my life when I had to make changes in production.
Is current browser version is sufficient for ensuring cross browser compatibility of your website?
People can have different type of browsers. browser versions, Operating systems and resolutions. If we calculate it will be around 2000 combinations of configuration. So I am asking to ensure cross browser compatibility across 2000 combinations.
I wouldn't even try to accommodate every browser / browser version combination out there. Your brain will simply explode!
Older browsers may not even support certain features that form an important part of your application. Even browser manufacturers don't support all versions of their software. At a certain point, they end of life them.
The client has a responsibility to upgrade his or her browser at some point, and that is as true of an individual consumer as it is of a business client. No-one can expect everyone to support their technology stack from now until the end of time. You know that is true ... otherwise we would all still be using DOS and the browser question simply would not arise.
There's another question to consider. How many people are actually using the different versions of each browser? Are you really going to put as much development effort into supporting a browser that is used by 4 people who may not even become customers as you are into supporting a browser used by 1,000 people who most definitely are?
This link might help you to gain a little perspective:
https://caniuse.com/usage-table
As you can see, for each browser, the version(s) being most used are highlighted. Older versions are now barely used at all.
If I were you, I would develop according to the latest broadly supported standards, make sure your app works properly across the new browsers and simply tell other people "tested on x, y and z. If your browser is older than this, please upgrade at [provide links]".
Otherwise, the one thing I can promise you is that you will never launch a product and the practical difficulties of trying to support 2,000 or more combinations, most of which no-one will ever use, will just overwhelm you and cause you to throw in the towel.
You do not need the sort of people who stubbornly refuse to keep up with the times as your customers. They will cost you far more than you will ever earn from them.
For the ones who are a little behind, chances are your app will work perfectly well. It is only the big changes in browser technology which ought to concern you. And there is nothing wrong with selling the idea of an uphgrade (they're free!) to customers who are slightly behind. You sell the idea on the fact that they can use features of your app (real or imagined) that their current browsers can't process. It is very common on websites to see links to the browser download pages of the different manufacturers. A few seconds' work on your clients' part will save you years of development frustration.
Don't be scared to tell your customers what you need from them. Most will be happy to oblige once they understand what is being asked.
Thanks for your reply@thomasm1964
This article will help understand why we need Cross-browser testing.
https://www.lambdatest.com/blog/what-is-cross-browser-compatibility-and-why-we-need-it/
I have to agree with @thomasm1964, Chasing after absolute perfection often leaves us at a stand still. Focus on a realistic set of parameters in a specific age-range and ship 😀
True.
I'm not denying that you need cross-browser compatibility. What I am saying is that you do not need to be compatible with every browser that ever existed. That is simply an impossible thing to do and will only become more impossible as browsers, client-side languages and CSS continue to evolve and screen resolutions / sizes change.
What you have to do is decide upon a sensible set of compatibilities and work to that set.
Anything which is not being used by a significant number of potential users should not be supported. It is as simple and as brutal as that.
If people are not using specific browser / screen combinations, how much business are you going to lose by not supporting that combination - and how much will it cost you to continue to support it?
Don't be frightened to set out your own boundaries in accordance with your own real requirements.
Hey! Are you talking about exploratory tests (ensuring that new features work across browsers and generally clicking around), or regression tests (ensuring you didn’t break old features with each change)?
Yes you got it right. All the functional testing which includes exploratory and regression.