Ok, i see what you were saying.

One thing that has come to mind while browsing the download options here and writing that previous post is that AR, as the apex of convergence of both hardware and software design has incredible potential, but equally incredible barriers. Think about all the little apps that are just great at one thing--there are quite a few. Then, as the functionality becomes more complex and there are more combinations of ways to do things, it's harder and harder to put together an app to do things in a slick and user-friendly way.

By the time you are capturing video, doing OCR and object recognition on the stream, manipulating the data, creating virtual objects to correspond with it, positioning those objects in the 3D space of the video image, and projecting the image in a way the user can interact with, and providing a useful UI to do so, there are multiples of world-class functionality all going on in a real-time, seamless way. That's a lot of complexity for any one developer or team to tackle.

All of that digression is a way of saying that we can start the Maemo AR revolution in stages, even right now, by building up the blocks and refining them. When the hardware is there, the optimized components can be integrated.

Make your apps modular!