Jordan Reiter wrote:
>...
>Well, one problem is that essentially Dublin Core shouldn't be made to
>embrace everyone's individual formats and schemes (schemae? schemi?)
>because Dublin Core is in its own right its own sort of scheme, right? So
>why don't people just list those types as proprietary. I mean, I don't
>know why someone would put
>
> NAME="DC.Resource-Type" CONTENT="This content not supported by DC"
>
>when they could just do
>
> NAME="MyProprietarySys.Resource-Type" CONTENT="This content supported by
> MyProprietarySys"
>
>instead. After all, if there going to use the Dublic Core Metadata, they
>should use Dublin Core metadata, right?
Not necessarily. My understanding is that, if you have some data for
an element that cannot be expressed in a popular scheme, then you should
put the data into the element anyway. As a fallback option, search engines
will just do string matching on the element values.
(This seems to be the assumption in the 4th Dublin Core Metadata Workshop
Report <http://www.dlib.org/dlib/june97/metadata/06weibel.html>)
It is also more efficient for search engines to index a defined set of
metadata elements rather than index any string that they happen to find
after "NAME=" on a web page. So the second example is likely not to get
indexed. If the number of element names gets large, it also gets very
difficult for users to formulate a query easily.
By the way, to avoid confusion, please use the agreed values in
examples (see <http://purl.org/metadata/dublin_core_elements>).
That is, use "DC.type" not "DC.Resource-Type".
Charles
------------------------------------------------------------------------
Any views expressed in this message are those of the individual sender,
except where the sender specifically states them to be the views of
Reuters Ltd.
|