[uf-discuss] re: HTML5 support

[Martin McEvoy]

  On 20/07/2010 03:57, Oli Studholme wrote:
> Hey Scott,
>
> On Tue, Jul 20, 2010 at 9:34 AM, Scott Reynen<scott at randomchaos.com>  wrote:
>
>> Making specific cases easier is the whole point of microformats, but it's not at all the point of microdata.
> “Making specific cases easier is the whole point of the class
> attribute, but it's not at all the point of microdata”
>
> Microdata — and semantic class names plus posh coding patterns for
> current microformats — are the method; a means to an end. Microdata
> vocabularies use microdata to express semantics, just as microformats
> use the class attribute etc to express semantics. Microformats are a
> little more concise in general (cough, datetimes ;-) compared to the
> same vocabulary in microdata (@class is shorter than @itemprop by 4
> characters, @property is optional whereas @itemtype is required etc),
> but the differences are not so great, and any class-based microformat
> can be written using microdata.

Im sorry but you cannot express *microformats* in microdata if you do, 
its cute, but It isn't a microformat because microformats *only* use  
class names, and a few choice rel-values.  If you move a microformat 
away from @class its no longer a microformat and shouldn't be described 
as such (we are a bit fussy about that :P).

This is why when someone starts talking about a "new microformats" or 
"microformats done better" the first thing I ask myself is "does it use 
semantic class names?" ... no well its not a new microformat or 
microformats done better.

Well the *good* news is HTML5 already supports microformats without 
adding any attributes at all (Yay!) .... that is until someone marks 
@class as obsolete!! ... joke.

Best wishes.

-- 
Martin McEvoy