A small update on what is going. See also Interoperability. Since Phil wrote a few testcases for Atom 1.0 documents other people followed and in general it seems that we can conclude that having a RFC number does not help getting it actually implemented correctly. The advantage here is of course that there is something to test against, where as with RSS you just have to guess what is correct. Aristotle Pagaltzis wrote Who knows an XML document from a hole in the ground? Instead of feed the nodeName of the root element became a:feed where a: was bound to the Atom namespace. Fun. People tested and results are available. Given that RSS 2.0 in a way requires namespaces to be supported it is quite strange that feed readers get such simple things wrong... On the other hand, these days feed readers are essentially tag soup parsers that also need to take namespaces into account. In a way they might be harder to write than an HTML parser.
Henri Sivonen made a counterpost yesterday to point out that prefixes on the XHTML elements cause problems as well. How unexpected :-) Now all that is left is some testcases for HTML to see if feed readers handle SGML comments correctly. Oh, wait.
At the other end of the universum Mark Pilgrim battles Apple by giving them documentation despite the fact that he is more or less ignored. Seems like a good thing though. Especially if it gives us quotes like: I don’t think Apple is trying to “break to web” or anything so grandiose. It simply never occurred to them to do this any other way. They view the web as an annoyingly inadequate infrastructure on which to build their latest proprietary network, in much the same way that old-school web designers view HTML as an annoyingly inadequate page layout language. Learning how the web works and going with its natural flow? It simply never occurred to them.