On Thu, 10 Jul 1997, Roy Tennant wrote:
> There's also the issue of extensibility. Do we need to identify every last
> possibility (I hope not) or can we devise a scheme that will allow some
> local variation with a flag so that it could be ignored if the record is
> processed by a system that is unfamiliar with the type.
I can hear the screams of agony already, but I have to say this: as good
as Roy's list is (and I think it's terrific), the notion of resource type
terms being drawn from a single controlled list is very nervous-making.
I realize that's the way the element has been defined, and that there
would have to be an excellent reason to change anything at this point,
but please hear me out:
For data originating in DC, a controlled list is fine. However, there
has been a lot of discussion recently about the potential of DC as a
switching language, lingua franca, filter, or whatever you want to call
it, between diverse metadata formats which will very likely not share
the DC list of controlled resource types. It is, of course, true that
they could simply leave the element out, but the reality is more likely
to be that people will map into the Resource Type element whatever
terminology they carry in their local data.
Would it be acceptable to have the DC list be the default, but to allow
a scheme qualifier to be used to indicate when terms are drawn from other
vocabularies? Or have free-text be the default -- since there are no
guarantees in this world -- but allow both the DC list and other
controlled terminology to be expressed via the scheme qualifier?
This would gather like information in the same element (rather than
forcing non-conformant form/genre terms into the subject element), and
it would allow applications to distinguish between "true" DC resource
types and unvouched-for terms.
Just a suggestion. Am I out of my mind (as far as this particular topic
is concerned...)?
--Robin
Robin Wendler ........................ work (617) 495-3724
Office for Information Systems ....... fax (617) 495-0491
Harvard University Library ........... [log in to unmask]
Cambridge, MA, USA 02138 .............
|