I find the idea of non linear programming to be pretty interesting. I sort of do this in Lisp, I program in terms of specific functions, but once I'm in the function programming becomes linear. I can do this less effectively in object oriented languages such as Java, since I have to write a whole class at a time. Some might call non linear programming top down or bottom up programming, but really it's about interest point based programming. Sometimes the interest points are at the top or at the bottom, but more often they are at some intermediary point in the complex tree or graph describing how your program works.
So, a 2D layout like Excel offers much more freedom, but I wonder if sometimes you'll want a 3rd dimension? And, if that's the case, is it possible to have a programming space that supports an arbitrary N dimensions, or N-N dimensions, ad infinitum?
Sure, it's probably possible, as certain types of cellular automata are turing complete. However, I've not seen anything more complex than a logic gate--no one thinks it's practical.
So, a 2D layout like Excel offers much more freedom, but I wonder if sometimes you'll want a 3rd dimension? And, if that's the case, is it possible to have a programming space that supports an arbitrary N dimensions, or N-N dimensions, ad infinitum?