It still needs some polish work like converting labels to Id's when appropriate and catching errors otherwise. I'll do that stuff this week.