optimized proof