Hmm, I regularly used things like code injection and/or edit and continue in Visual Studio to dynamically change the UI code. There were limits, but it worked g

Hmm, I regularly used things like code injection and/or edit and continue in Visual Studio to dynamically change the UI code. There were limits, but it worked great for doing things like incrementally changing the positions of elements, turning on/off OpenGL states, and the like. If you made enough changes in a single run it would eventually give up and force you to recompile, but by and large it was pretty seamless. It was great for things that did not require reasoned thought, but required experimentation (what does OpenGL do if I set X to Y). All my work these days is algorithmic, so I am not entirely sure what state of the art is, but it couldn't have gotten worse, could it?