Great to hear your thoughts on cross-platform support. I just wanted to comment on:
the real value lies in your hardware product
vs.
Without the software the EyeX is simply yet another webcam
I agree that the hardware part is where most of the value is, but the way I see it there is a significant value also on the software side.
The hardware has been evolving for over a decade and is optimized for tracking eyes. It is not just a webcam, since it uses a combination of tailor-made hardware components, algorithms and control logic to be able to compensate for head movements, varying light conditions and different eye types. See this page for some info on the different parts of the system
http://www.tobii.com/en/eye-experience/oem/core-eye-tracking-tech/
(sorry about the marketing talk, it is aimed for OEM companies and not developers)
However, our experience from working with 3rd party developers and hardware integration partners for many years is that it is not enough with a good eye tracking hardware with a low-level API. To be able to develop a descent application, there is a lot of work with handling hardware connection and error codes, screen setup, gaze data filtering, object tracking, additional input handling and the actual user interactions. Most developers have given up way too soon. That’s why we have invested a lot in the EyeX Engine and EyeX SDK for Windows.
We would like to prove the use cases on Windows first, before we can move on to other platforms. It would be a possible option to open-source parts of the technology stack for other platforms and use the Windows solution as inspiration for the community, but we have not decided whether to take this path yet. Good to hear that we have some contributors out there, should that be the case.