>  If not dc, then who?  If we call it bib1, instead of dc, is it any
better?

Actually yes it is (though certainly we're not going to call it bib1).  The
point is, dc is an abberation, in the srw context.  We need dc, to support a
number of applications, OAI, for example.  So getting rid of dc altogether
is not an option.

We need a dc schema, that is. You know, I can't think of any compelling
reason why we need a dc index set, though.  I'm tempted to suggest we get
rid of it. Someone please remind me why we need it.

We do need a general purpose index set, and we get in trouble if we try to
force dc to be it. We can't have a dc schema and a dc index set unless the
two are tightly coupled, can we? And, as Eliot and others suggest, we want
our indexes to remain abstract, not tied to retrieval elements, don't we?

--Ray