On Tue, 6 Aug 1996, Stu Weibel wrote:
>
> Meta Folk,
>
> The issue of Scheme specification is one of those detail devils we here
> so much about. Its something we have to come to grips with, but its
> hard to balance the need for simplicity (to promote the sort of rapid
> adoption that is critical to achieving `market-presence') and the
> precision necessary to support machine parsability and a coherent data
> model.
A quick comment from the flat shoed perspective..... I think we need to
have a bit more context about how the schemes will be used, in
particular how they will be created, and if they are permitted then what
about control? I think there is a danger that the simplicity of DC will be
lost if qualifiers get out of control. Could we keep DC simple and use
Warwick framework to incorporate the richer metadata?
With the proposed schemes is it envisaged that authors or web
administrators will enter schemes? I guess not as they will not have the
knowledge of different schemes. It seems to me schemes are needed when you
try to squeeze a more complex record into DC which takes place further
down the line than HTML.
Rachel
-----------------------------------------------------------------
Rachel Heery, Metadata Projects Co-ordinator
UKOLN (UK Office for Library and Info Networking) tel: +44 (0)1225 826724
University of Bath, Bath BA2 7AY, UK fax: +44 (0)1225 826838
Information on ROADS: http://ukoln.bath.ac.uk/roads/
DESIRE:http://www.surfnet.nl/surfnet/projects/desire/
