That's a known bug that I've been meaning to look at. It's not just the slashdot feed but several others as well and it's quite annoying to me too. I'm using a third party library called SimplePie that is supposed to parse the RSS / atom text into usable data, but the library often bombs out instead of returning anything. It's a pain because the fatal error in that library causes the whole feed process to completely stop until cron restarts it again. This causes not only the offending feed to not get data but all other feeds afterwards to halt. Feed parsers are complicated because a lot of sites return malformed data with stray tags all over the place. In an ideal world you could just parse an article with a strict xml parser and everything work! But far too many sites forget to close their html tags or use some invalid unicode strings that cause the parse to fail.