Hi there, everybody on the list!
I have been following the debate on resource types and found it
extremely interesting and stimulating. And now, as a little side
discussion about DLOs has set in, I want to come up with a few
comments of my own.
> On Thu, 24 Jul 1997, Judith Pearce wrote:
>
Should Dublin Core be being used to describe these resource types in
the first place? The fact that they're not document-like objects
seems to me to be a key issue. Wasn't there an intent originally to
identify other object types that might be handled through a Dublin
Core-like structure? I see this as different from extending Dublin
Core to handle specific types of document-like object and having a
Resource Type hierarchy to categorise document-like objects.
> On Thu, 24 Jul 1997, Jon Knight wrote:
>
However I disagree with using DC to represent metadata for non-DLOs,
even if those non-DLOs have DLOs that talk about them. A GIS or a
genome database, etc doesn't sound much like a DLO to me and so
shouldn't have DC for it.
> On Thu, 24 Jul 1997, Jon Knight wrote:
>
DC was designed with document like objects (DLOs) in mind and so,
unsurprisingly, isn't do too good a job with non-DLOs.
> On Thu, 24 Jul 1997, Simon Cox wrote:
>
Fair enough. But I think we will find that the sub-types of
"Data" will evolve and grow - and then there is a danger
that "Data" could become the dumping ground for lots of
stuff that doesn't seem to fit comfortably anywhere else.
"Data" could become the "Misc" that we thought we were
avoiding.
Listening to that, it occurs to me that perhaps some further
considerations need to be made here. When you look for instance at
Andrew Daviel's DUBLIN CORE TYPEs, then you see many non-DLOs listed
there. But, according to Jon Knight, non-DLOs are outside DC territory.
There obviously is some tension that should be reconciled in order to
make DC operable and secure it as the means it is supposed to be.
I think the effective list of DC resource types will eventually
depend on our understanding of DLOs or the range of documents we
claim to be covered by DC. We need to be clear (perhaps clearer than
we are at the moment) about the basic unit(s) that we want to
describe, and from that definition the legitimate resource types can
then be derived. I believe that resource types and DLOs (or whatever
we replace them with) are interdependent and need to be thoroughly
treated that way.
What follows from this is that we may be back at one of the original
questions in DC history (hey, is it really that easy to have "a
history"). Is it really only DLOs that DC wants to address and cover?
As far as I understand, this was an isssue at DC3 (the CNI/OCLC Image
Metadata Workshop), and it was decided (or should I say: the opinion
prevailed) that DC indeed should have the potential to go beyond that
range.
But the more you increase your territory, the more you are in danger
of falling victim to what the historian Paul Kennedy has aptly
called "imperial overstretch." In order to avoid that trap, it seems
to be advisable to get a little clearer about our own ambitions and
to arrive at a mutual understanding of what we want to cover and what
we want to leave out. Is DC bound to document-like object or might it
as well go after all conceivable incarnations of electronic data
streams now and in the future (as Simon Cox seems to suggest)?
Ralf
********************************************************
Dr. Ralf Schimmer
Niedersaechsische Staats- und
Universitaetsbibliothek Goettingen
Platz der Goettinger Sieben 1
37073 Goettingen
Tel: +49-(0)551-39-5230
Fax: +49-(0)551-39-3199
e-mail: [log in to unmask]
********************************************************
|