I’d like to get into more detail, when it comes to my ideas on visual programming, after a chat with Dane of Colorado. There are a lot of details to be worked out, naturally, especially with the assertions I made in my last post.
First things first, gestures and motion won’t solve everything – there’s a lot to be worked out. You need primitives.
You need primitive data structures, and you need primitive “actions”.
For example, adding two numbers. If you “grab” two numbers in VR, and “join” them, should that add them?
Well, that depends on how you define your programming environment.
I believe there will be many, many programming environments that are specialized to their use cases. For example, I’m building Vonnegut, a visual programming environment for interactive fiction. This cannot be used for a blog, or a commerce site.
If you did want to build those sites, though, there should be primitives that you can use. For example, “add to cart” should be a common “sign” or gesture you can program to a given motion.
So the programming happens on many levels. If it’s basic, it can happen via a simple gesture. If it’s complex, it can be programmed to a gesture, then used commonly as a gesture. And if it’s super complicated beyond the basics, you can always fall back to good ol' text programming to create a common gesture that can be used in these ways.
I think that’s the key. These gestures have to be programmable, widely usable, accessible…that’s all I’ve got for now, but stay tuned!