|
Validation may be difficult, but how important is it in typical microformat scenarios on the Web? I appreciate that this is machine-readable data we're talking about, so liberal browser-like display may not be an option. But some client liberality will almost certainly have to be considered, but hopefully not to the extent of RSS (and how many aggregators have validation in their pipeline?).
For my own applications I plan to pre-clean pages with Tidy before ahead of XSLT (to RDF/XML, as Daniel suggests). I anticipate a proportion of junk data, but will cross that bridge when I get to it.
|