I’m pretty convinced that to make effective use of touch (and your fingers, if they are anything like mine) we’ll need new interface paradigm altogether. I’m not very precise with my fingers and a lot of existing applications (and UI themes) require precise positioning to click on tightly-grouped buttons and skinny widgets. Go ahead and try to pretend to use your fingers to interact with the form you get with “POST REPLY” feature of this forum and you’ll see what I mean
I agree we can’t throw away the existing application base, but on the other hand there are other applications we can explore. I’m personally not motivated to get plain old X11 apps running on my table, but much more motivated to get new apps written for it ... what do I have in mind?
Trek-style comm link: by this I mean the ability to talk to anyone quickly, no matter where he/she is located. Admittedly this has very little to do with touch and a lot more to do with locating technology, but still, the touch interface is the make/break factor. I want to use touch interface to establish a link with a person in under five seconds.
Family photo album app: yes, there is a bunch of photo apps floating around, but when you have thousands of pictures organized in folders, there has to be a better way to look at them than randomly throwing a pile on the table ... that’s not how you look at your photos IRL anyway.
Kid-friendly games and puzzles: OK, I’m not looking to replace all actual toys with my table, but there are some toys that I think would benefit. If your kids are like mine, they lose jigsaw puzzle pieces often, and they hate turning the pieces face-up first
I’d like to code one up that would actually be challenging, and sane to work with. I want one that will work with 5000 pieces, for example, without throwing ALL 5000 pieces on the workspace at the same time (that’s where the sane part comes in).
Environmental simulation: I’d like to simulate marine, desert, and marshland environments on my table. We are an active family and I believe touch-based interface has a lot of value in this area. I’m thinking of using something like Ogre3D to help with this. I know there are many simulators out there such as http://www.aridocean.com/ but I’m not aware of one that is distributed under open-source license, so I’m planning on writing one