Instead of defining a search space, the user provides a function that can randomly give you a valid point in input space
Rocketsled picks, say, 1000 random points using the provided function and runs the one that is the "best" according to the current surrogate model and selection strategy. Since the search space has continuous dimensions we don't have to worry about duplication
Reason this is useful is some problems have complex algebraic constraints that cannot be known a priori. Also this just makes the bounding of the problem a little easier. https://groups.google.com/forum/#!topic/fireworkflows/O5x6CCMHjqo
Perhaps restrict this to float dimensions only such that no duplicate checking problems occur