It's amazing how regularly people reinvent this. Truly the perennial failed paradigm. I once worked with a programmer who had a bee in his bonnet about visual

It's amazing how regularly people reinvent this. Truly the perennial failed paradigm. I once worked with a programmer who had a bee in his bonnet about visual modeling languages. Every week or so he would arrive for another hour-long harangue on the topic. Nothing I said could convince him of anything, until one day it occurred to me to suggest that he take a moderately complex real-world model (from one of our clients) and mock it up in his visual language. He disappeared for three months, then showed up at my office, said "You are right! Can't be done," and left. That was the end of that. I was lucky, though, because this guy was smart. Had he come back with visual spaghetti instead I would have been done for. I should add, though, that people have told me that visual programming languages have found a niche in industrial process control. I have yet to hear a good answer as to what it is about this domain that makes visual programming actually work, since it's so infamous for being useless in pretty much every other area of computing.