Is this the right way to manually test?
I'm new to software testing, been working in the field (manual tester) for 3 months.
I test a product that the end user accesses through a web-browser, we support Firefox, Chrome, IE11 and Safari.
And we support Windows 7, 8, 8.1, 10..and so on
Yet every test we run (my team) is from our work PCs, that run Windows 7 64-Bit Enterprise.
Does that seem right to you?
There might be defects that are unique to:
IE11+Win8+Our web service
IE11+Win8.1+Our web service
IE11+Win10+Our web service
Firefox+Win8+Our web service
Firefox+Win8.1+Our web service
Firefox+Win10+Our web service
That we're missing, isn't that so? :S
On running tests on different versions / platform / etc...
* you can use VMs (ex: VM ESX Server running on a Mac Pro can get you a wide range of platforms)
* Using a browser as a service - SauceLabs, BrowserStack, etc.. have bunch of browsers setup for you, which you can connect to using a web browser
* Using Platform as a Service, like Microsoft Azure to launch VMs of any windows (from 7 and up), then using a Mac with parrallels to test versions of OSX.
On what you should test based on browser versions....
The older the HTML version used, the less differences you'll see in browsers.
The older CSS version is used, the less differences you'll see
For the change being made, you'll want to look it up in Quirks mode, Compatibility overview, to see if the underlying technology has differences. Pay particular attention to Red entries, as those require extra "Polyfills" or special alternate code branch renderings to make backwards compatible.
2. Are native plugins used. (ex: Flash, Silverlight)
Because native plugins run not in the browser, a separate process, you need to test those across different platforms. In this case you don't have to separately test IE11 vs IE8, but you do want to test Windows vs Mac, and Windows Server vs Windows Desktop. As they have differences in the underlying platform.