Help me understand what your rationale is. Are you saying that SQL + source control is good enough? Do you think it's a fools errand to build tools like this which add additional abstraction layers for schema revision?