There is currently a lot of activity in the area of Microformats - the idea behind it: to store information blocks not in XML, but in predefined HTML. CSS classes are then used to define what a single format is. Logically, there are problems with colliding styles - what a surprise. I myself am always amazed at how much energy developers can spend on stupid ideas.
We once had HTML that not only dealt with semantics but also with layout. And that produced the all-time favorite FONT-TAG orgies on HTML pages. Over time, most people have come to the realization that separating semantics and layout makes sense - semantics as a basis for marking up content, layout in the CSS files, and as a connection between these, the IDs and classes on tags. Additionally, with DIV and SPAN "anonymous" tags without predefined semantics (except "this is a block of text" and "this is an inline stretch of text" - where this meaning can be easily overloaded), for things that don't work with the normal semantic markup (which is mainly due to the rather stupid idea of HTML that there are markings for headings, but no markings for sections of text to which these headings would belong).
What do Microformats do now? Well, the same stupid idea of misusing something - namely in this case the connecting pieces between semantics and layout mentioned above. Microformats give these a meaning - for example, a DIV with a class 'description' would then be the description of a review - read the details in the hReview reference. Sorry, but this must inevitably lead to conflicts - have the idiots never heard of namespaces? The Microformats explicitly address XHTML - and that has exactly the purpose of embedding namespaces. And if you think you have to implement such a stupid idea - couldn't you at least be smart enough to give the parts more cryptic but unambiguous classes?
As I said, it's amazing how much energy goes into such stupid ideas that are doomed to create more problems than solutions.