[uf-discuss] A Basic structure for Semantic-Data ??

[LucaP]

Think of any web context where listings of physical /or financial
quantities is crucial -take an
earthquake event as an example; 'hcalendar' with embedded 'geo'
perfectly encodes 'time&place' information; but 'magnitude' and
'depth' would go in the <p class="description"> </p> part and they
should retain full parsability by remote data-processing applications.

Reusing the hcard pattern for telephone and rel-tag, something as simple as:
	
	<span class="data-value">
	   <a href="/depth" rel="tag" class="data-name">Depth</a>:
	     ( <span class="data-value">21.7</span> +/-
		  <span class="data-error"> 0.2</span> )
		  <abbr class="unit-measure" title="km">Kilometers</abbr>.
	</span>
	
	<span class="data-value">
	   <a href="http://en.wikipedia.org/wiki/Richter_magnitude_scale"
rel="tag" class="data-name">Magnitude</a>:
	     <span class="data-value">5.8</span> .
	</span>
	
would suffice. Easy an general enough to cause wide-spread
microformats adoption outside web-geek circles!
here's another example from a weather site:

	<span class="data-value">
	   <a href="/temperature" rel="tag" class="data-name">Temp.</a>:
	     <span class="data-value">118</span>
		 <abbr class="unit-measure" title="°F">Fahrenheit</abbr>
	</span>


Trivial considerations:

	-data-error and unit-measure would be optional, since the former is
needed only in a scientific context, while the latter is not defined
for scalar quantities (i.e. the Richter scale for earthquakes
magnitude above)
	
	-include-pattern should be supported, very handy to avoid html bloat
in data-tables
	
	- since there are internationally recognized System of Units in place
since the 60s, 'unit-measure' should conform with them, and thus
restrict the set of allowed values for the title attribute to those
specified in http://en.wikipedia.org/wiki/SI and similar documents for
financial quantities (currencies) (...)


I cannot find any previous discussion about such a fundamental
data structure, aside from XOXO, unit-measures and measurement-errors
need ad-hoc classes and rules; I thus believe it would be worth
discussing a new Elemental microformat (??)

Ciao!