I heard that the “Just” Use HTTP session by Sam Ruby was good (it was the first day of ETech, which I missed). Not to fear! The slides for the session are online. They talk about a bunch of issues that go in and around trying to use HTTP - character encodings, URL encodings, defaults, case of specifiers, where markup is valid in documents. Working with RSS one quickly becomes intimately familiar with the differences between XML processing as practiced between components of your own application and processing of documents out on the Internet. Notice I don’t say XML processing out on the Internet. Apparently there’s no such thing. So many documents out there are nonvalidating so often that it’s not even worth trying to process them with a standard XML parser.