Hello folks!
I’ve been working on my wxPython project for about two years. It’s high time for me to do some automated GUI testing.
I already have a nose test suite for the non-GUI parts of my program. It’s pretty comprehensive and it works great. But I must test the GUI as well.
Let me try to describe my ideal continuous-testing setup:
I want real, mouse-keyboard-screen testing for my GUI, not just manually posting events. And the tests should work cross-platform.
Also I want to start using Hudson, and have it run my existing test suite plus the GUI tests. Every time I push a commit of my project to GitHub, I want Hudson to pull it and run all the tests, including the mouse-keyboard-screen GUI tests, on all three platforms, without any human intervention. (Probably using a bunch of slave machines, physical or virtual, that run the different OSes.) Then I want it to post the results on one single page where I can review them.
I may not get all of the above, but I want to get my concept of the ideal system right before I see what actually exists and try to get a system that’s as close as possible to my dream system.
My questions:
-
Do you have any other additions/suggestions to this dream continuous-testing setup that might make the developer’s life easier?
-
Do you know if what I want is possible? If so, how?
-
Do you have your own continuous-testing setup for your wxPython project? If so, what is it like and what tools are you using?
Best Wishes,
Ram.