A forest-fire model and some thoughts on turbulence