This post discusses the challenges sysadmins face with feed readers and crawlers. It highlights the neglect of HTTP meta-headers, like Accept-Language, and the issues surrounding ETags and Last-Modified headers for optimizing data retrieval. There is a strong sentiment towards the inefficiency of current polling methods and the importance of utilizing tools like WebSub to improve the publish-subscribe model for RSS feeds, reducing unnecessary data requests. Various comments emphasize the need for better practices from both feed creators and readers, such as respecting TTL values and adjusting fetching strategies to mitigate bandwidth waste. The conversation underscores ongoing frustrations with misbehaving crawlers and the user experience implications of poorly implemented feeds.