Mikael Nilsson wrote:
>
>
> "Machine-processable information about a resource, with a
> machine-processable reference to the resource."
>
Or, in other words, in order to define metadata, it is less interesting
to specify *what* we describe using metadata, and more interesting *how*
that description is made. This is a radical departure from previous
discussions on this list.
/Mikael
--
Plus ça change, plus c'est la même chose
|