I deployed my own RSS reader to handle sites that update infrequently. That's still ultimately a pull, I just have an agent doing the pulling, and then I visit the site for my reader.