There's a lot of solid GUI automation python packages but I have a specific set of requirements needed for a recommendation to one of our more advanced QAs.
I'd like to know in 2019 what is the easiest to pick up and run with python GUI automation package that supports image-recognition instead as well as coordinates being hard-coded for finding and interacting with GUI elements.
There's no requirement to scan/assess under-the-hood GUI ids or handles.
The end-goal is to integrate it with Appium for automated user-workflow-style testing.
Currently it's only needed for Windows but I suspect that will change later. I will however accept Windows-only answers.
Update: One more thing to mention: support has to be good for it, whether that's an active Q&A community or in-depth how-tos. That makes SIKULI a no-go from my own personal past experiences with the product. None of my attempts to get feedback for highly specific questions from them produced any feedback, so I can't in good consciousness recommend it even if it is ostensibly a very easy to start using GUI scripting tool that supports image recognition.