Code: managed to slot in 90 mins on my main project last night, after I'd put Cordi and Miles to bed, and before Toni and Edmund got back from the ballet. They saw the Nutcracker at Sadler's Wells. I hadn't done any work on the code since late November, so it took a while to get up to speed. I seem to be homing in on a core issue for levitt's object system: the relationship between persistent object state and the run time stack frame. Specifically, how object properties can move from local state onto the stack. I hope I'm going to get a chance to work on this over the holiday. I can feel the beginnings of a solution starting to fight its way out of the back of head.