Rhetorical Question
Dare Obasanjo: How come those errors weren’t pointed out by the validator at http://www.feedvalidator.org
The Feed Validator hasn’t supported Atom 0.3 for quite some time, and it clearly says so. But I’m assuming that you know that, and you are asking why. The reason is quite similar to the reason why Atom 0.2 and Atom 0.1 aren’t supported.
Truth be told, the Feed Validator doesn’t really distinguish between the various versions of RSS. Try validating an RSS 0.92 feed that has neither a title or a description. Or an RSS 0.91 feed with namespaced extensions or a guid. There are some minor exceptions, in particular RSS 1.0 feeds are not only validated as RSS but also passed through an RDF parser.
For a period of time, I tried to support both Atom 0.3 and Atom 1.0, but not only was the code a mess, what ended up happening is that the code wouldn’t report on errors where Atom 1.0-isms were being used in an Atom 0.3 context — an example of this would be type="html" when found in an Atom 0.3 feeds like flickr’s would not be flagged as an error.
Additionally, the combination of type and mode is poorly understood. Take a look at a typical WordPress feed in Firefox 2.0. The summary is type="text/plain" mode="escaped", but what does that really mean? Assuming that the feed is read by an XML parser, I can read the Atom 0.3 spec in a way that indicates that WordPress is right and Firefox is wrong. I can also make the case that Firefox is right, and WordPress is wrong. What’s the right answer? Atom 1.0 for WordPress, of course.
This is a situation where some have complained, but few have volunteered. That’s why whenever anybody has a need for an RSS version specific check, I point them at Validome. They even check for the differences between UserLand 0.91 and Netscape 0.91.