This reminds me of a project I worked on right after college, doing the initial implementation and research of something now called FLX (https://books.google.co

This reminds me of a project I worked on right after college, doing the initial implementation and research of something now called FLX (https://books.google.com/books?id=DiOr9P2oA6YC&pg=PA115&lpg=PA115&dq=feature+language+flx&source=bl&ots=wLZG_mHcj9&sig=ACfU3U0mJBnlsBzUyGK6X-R1-sf0IutIgA&hl=en&sa=X&ved=2ahUKEwiq-emc7IHnAhXVHjQIHT9eAP0Q6AEwAXoECAkQAQ#v=onepage&q=feature%20language%20flx&f=false https://books.google.com/books?id=DiOr9P2oA6YC&pg=PA115&lpg=...) One of the issues when then number of modules grow to a large number is that they may all have different conditions to keep track of. It’s then possible that the modules all contradict each other and never yield an executable state. Imagine trying to debug a system like that. The next layer is then to ask, with all the modules and their output state, will their combinations ever be satisfiable. At that point, we turn the states into a large predicate logic equation and can use a SAT solver to figure that out. I left off there in 2005, but the project is still being researched on. If you’re interested I can get you in contact with the professor.