Hi all, I am kind off new to automated testing. I have a lot of experience in testing itself, but we want to automate so processes for our desktop application, at least for core functionality, so I’ve been searching for libraries to use with robot framework but I am getting a bit confused here.
There seem to be a few which support test automation of desktop applications (windows) but most of them are deprecated or just not actively updated anymore. Like whitelibrary, RPA library etc.
So my question is, what would be the best library to use currently? Which one works well and is still maintained till this day?
Or maybe I should look for something completely different here, I don’t know. I hope someone can help me out!
There is also PlatynUI Library which is currently in pre-release but I heard should be released soon, This library should be a serious consideration if you need to test the same app on multiple OS’s or you need to test an app on an OS that’s not windows (i.e. MacOS or Linux)
Start with either FlaUI or PlatynUI, as If these don’t work there are other options, but they all get harder from here.
What damies said but also worth to point out that things still depend on what technology the app was created with. QT has its own stuff (squish which they bought from Froglogic) and if the app is electron, Selenium is still good option ..
sorry for the late response, I did see it pop up in my email but forgot to write back. Very useful feedback, thank you. I did hear and read about FlaUI indeed, I will do some research on the usability of the library.
Well, that’s another challenge, the app was build with VS, DevOps, C# but the buttons, textboxes, fields etc were never decently tagged or named. so it will be a project where developers are also involved to get the app ready for automated testing, which is currently not the case.
They did try to get it working with expensive automate tools in the past, before I started working there, which did not work as well as hoped for. So they left it as is, but I would like to get it up and running somewhere next year. At least for the core functionality. It is a lot of work to manually test everything and nearly impossible to test every possible scenario.