Although it appears that the code for both of these pages is correct, it's actually not. The bottom one throws an
error on the server at the first try, then tries another path and finds the image. The end result is that the page is
not broken, but it wastes system resources.