Lambda the Ultimate

Prograph, mTropolis, Director... none of them ever works out in the end. I don't think that's an accident.

I'd like to see a survey of paradigms that might be suitable for the sort of work ANSWER is attempting to answer. Here is what I have:

Constraint Programming

Describe some general limits and some heuristics, and allow the machine to provide a solution meeting the needs of the designer. With enough

Physics-based / Rules-based

Describe some initial relationships and rules for how relationships and rules change over time. Introduce interactive input into the model as a relationship that changes over time outside local programmatic control. Compute states over time. Has been implemented in many paradigms, though the denotational FRP implementations seem cleaner than most.

Focus on Domain Specific Languages

Ensure the artist can express his ideas clearly and concisely in a DSL. Not really a paradigm, but important to those 'initial conditions' and 'constraints' and such. It has been observed that artists almost always need to escape the optimized syntax to handle the corner cases, so an embedded DSL - with access to more power when critical - might be the wise option (lest you end up reinventing a GPPL).

That's all I have, though I could probably name more variations on the rules-based approaches.

I'm thinking about the tool I would love for my story. It's a story about technology and innovation, shouldn't I tell it using technology and innovation?

I'm not thinking about rules at all, but more examining the kinds of "things" I need to show. Perhaps also looking at universal iconography to help bring the "rules" into focus.

What's happening with maps right now is a perfect example. Almost anyone can now throw up a map and post stories. And have others contribute. Their story relationship is inherent in the map.