Let me start by saying that I posted this same inquiry on the Macintosh forum -- but haven't gotten any responses. Thought I'd see if the traffic was higher here and if anyone had any thoughts. -- bkr

Seems to be about the only tool that anyone is mentioning these days for Mac automation.

At first blush, the website would seem to imply that they use bitmap / screen coordinates as the basis of their object recognition / interaction (... drag a box around the object you want to interact with then give it a name ...) but maybe they have a "better" implementation.

Has anyone done anything substantial with the tool beyond an eval? If so, how did it turn out?

I come from a long-lived QAP/Silk background and am frankly sold on the theory of their approach (application framework, object orientation, etc.), so comparative analysis would also help.