Hey Pete

I hope you're not suggesting dc:date and dc:coverage have overlapping
semantics?

Andrew

-----Original Message-----
From: A mailing list for the Dublin Core Metadata Initiative's Usage
Board [mailto:[log in to unmask]]On Behalf Of Pete Johnston
Sent: 05 September 2005 14:33
To: [log in to unmask]
Subject: Re: Something to look at re: Coverage


Hi Diane,

I must admit I think that document oversimplifies things, because it
glosses over the fact that rather than having "geographic place or
location field(s)", metadata formats may have _multiple_ "fields" to
express _different_ types of relationship between the described resource
and a place/location.

e.g. as we've already discussed, "aboutness" and
"relevance/applicability" relationships are different; a format might
also deploy different fields to distinguish "place of origin", "place of
access" etc etc etc.

So the use of a particular "field" depends on whether it is appropriate
for that _particular_ type of relationship between the described
resource and the place.

If an application is using dc:coverage as a "geographic place bucket" -
i.e. to capture a relationship of _any_ type - between the described
resource and a place or location - then (it seems to me) that
application is really mis-using the dc:coverage property: even given the
current permissive definition of dc:coverage, it is _not_ a
"geographical bucket".

And the "it's just a bucket" approach becomes even worse if you take
into account that dc:coverage has a temporal aspect too: if it is a
"temporal bucket", then values may include any date or period associated
in any way with the resource - just like dc:date :-(

Pete

-----Original Message-----
From: A mailing list for the Dublin Core Metadata Initiative's Usage
Board [mailto:[log in to unmask]] On Behalf Of Diane I. Hillmann
Sent: 02 September 2005 22:35
To: [log in to unmask]
Subject: Something to look at re: Coverage


Folks:


You might want to look at:
http://oai-best.comm.nsdl.org/cgi-bin/wiki.pl?GeographicPlaces


as an example of how implementors are recommending use of Coverage.  As
I mentioned earlier in the conversation, in most situations the
recommendation is to use Coverage as a geographic place bucket, without
reference to "aboutness."


Diane