I disagree on two counts. Firstly, in terms of the vision, the biggest problem, that of how to manage naming of things (and relationships between them) has already been solved: URIs. The existing web is a globally scalable framework, adding better data support isn’t such a big step given that.

Secondly the approach to data integration RDF offers is finding a lot of use *off* the Web (and hence a long way from the vision). The two main contracts I’m working on right now are for folks that have turned to SemWeb technologies because traditional approaches were found to be wanting for certain requirements (particularly for handling “incomplete” information, that doesn’t fit neatly into the usual RDBMS or XML paradigms). Compatibility with the Web is relatively low on their priorities.

But on the basic point, I don’t really see any problematic conflict between microformats and RDFa, they are both basically just ways of using content formats to carry explicit data (and remember microformats can quite easily be interpreted as RDF). If there is any conflict around those parts it’s between HTML and XHTML 2.0 (and HTML 5).

Ok, so I guess I’d define microcontent as small pieces of human-readable content associated with a significant amount of metatadata.

In that context, personally I can’t see myself using RDFa in the near future because of the format/content issues. In practice microformats and/or embedded RDF and/or separate content & metadata seem to cover the requirements better. That is at least until RDFa plays nicely with XHTML 1.0.

For non-(micro-)content data on the Web, i.e. stuff that isn’t intended at least in part for direct human consumption, I haven’t see any viable alternative to RDF.