The user will  never see the context set and will never see the record
schema because the client will do the mapping from title to dc.title and
from tag 245 to title. So what is the issue?

Theo


>>> [log in to unmask] 31-8-04 11:49:10 >>>
At 09:36 AM 8/30/2004, Ray wrote:
>I agree of course, but as there isn't anything that formally ties the
DC
>index set to the DC schema, I'm trying to imagine how such a
recommendation
>is to be formulated. [...] Or a more general recommendation to apply
to
>all such pairs? In that case how do we assert this relationship?

It is my belief that the root problem with DC elements is that they
try
at once to be abstract (as indexes) and concrete (as record schema
elements).

Obviously, it is quite useful to have a small set of named concepts
(title,
author, subject, date...) as abstract search access points. These
concepts
could be asserted in the cql context set, along with serverChoice and
anywhere, and so would be readily available to other context sets. The
DC context set can, of course, have its own elements for those cases
where
the content actually has a concrete DC structure.