> I'm not sure why this takes into account the right endpoint of the > extent on the left but not the left endpoint of the extent on the > right. I think it might just be assuming that all break-aligned > grobs have an extent starting at 0.
Or maybe it's simply a thinko...
Werner
