I'm working with my team on developing a web app, and our users are expected to come from across browsers (Chrome, Firefox, etc) / OSs (Windows, Mac, Linux) / devices (Mobile, Tabs, Desktops).

We are now working to define our testing plan, and we feel that trying out each and every combination (e.g. Chrome - iOS - iPhone, then Chrome - iOS - iPad, etc) is going to be overkill. Also reading through: https://stackoverflow.com/questions/9853497/any-difference-between-chrome-on-mac-and-chrome-on-windows? & https://stackoverflow.com/questions/4299799/cross-browser-compatibility-issues ; we get the view that there are sufficient similarities across platforms for there to exist a smaller sub-set of combinations, which should take us to (at least) 90% reliability.

My question is if someone has done this exercise before, and what their conclusion was - i.e. which browser / OS combos did you finalise on? Else, if that is always going to be very specific, what's the best way for us to figure the right set for us?


My conclusion: Test all browsers combo's, only plan better.

I prefer working in Firefox. Firefox is in my experience fairly up to date, without having to get all hacky to make things fix. When I'm (almost) done and it works in Firefox, it often does in Chrome and the latest IE's.

After I've done this, I test on mobiles. Android's default browser, Chrome for android and iOS devices. I do these at the same time as a fix for one, might break it in the other. These are more fragile browsers and should be treated accordingly.

If I know I have to build a site that has to work on every device, I dail down on the cutting edge tricks. Those are fun, but a lot of work to make crossbrowser. Fundamental functionallity may not really one that, stick to time-tested solutions.
This is where experience comes in, after some years of working in this area, you can assess whether or not you can use something (or you'll know where and how to check).

Check you statistics and decide from there. You should check your statistics and see what browsers your visitors use. Those should be tested the most intense. The browsers with less users should be tested aswell, but if you have a time restriction you can test the important functionallity the most and click through the other elements a bit faster.

Not the answer you're looking for? Browse other questions tagged or ask your own question.