Re: RDFa vs Microdata, and the separation of data and presentation

>
> However, RDFa, Microdata and Microformats, _are_ about giving the actual
> structure of the document semantics.
>

I know, and I'm suggesting that this introduces far more cost than benefit.


>  So, for example, when I have a block of HTML that describes a calendar
> event, I can give it meaning by using appropriate markup.
>

Which you can do in my scheme, too. The ID linkage between the HTML and the
DATA are important, I agree. It's not enough to have a page and a dataset
with no clue how they connect. But I don't see any justification for
encoding the "event-location" *predicate*, for example, into the DIV that
displays the location. What is anybody going to do with that information
that wouldn't be much more straightforward with just data ids on the HTML
elements and all the rest of the data structure in JSON?