Ray Denenberg writes:
 > > the proposed DC+RMD schema is potentially useful, but it doesn't
 > > solve the problem that this thread is addressing.
 > 
 > To be clear: the solution outlined in my first message doesn't
 > solve the problem, but are you saying that the amended proposal
 > also doesn't?
 > 
 > That is, to add an additional 'any' element to the schema for an
 > additional namespace, this second namespace for descriptive data
 > where we can put useful elements, one of which would be 'price'.

OK, sorry, I'll try to be more explicit.

The (immediate) problem on the table is how to convey price
information in an SRU record.

Solution 1 is to use a schema that includes that information, such as
ONIX.  That is a good solution, providing both syntactic and semantic
interoperability.

But some people find that too heavyweight, so the proposed solution 2
is to use "DC extended", in which DC records can be augmented by _any_
additional elements not in the DC namespace.  That's what Theo does.
This approach does solve the syntactic problem, in that it provides a
means for a server to include price information -- for example, as
	<x:priceInDollars xmlns:x="some.uri.">2.99L</priceInDollars>
but it does nothing to solve the semantic problem because an
out-of-band agreement is required for the client to understand the
meaning of this element.

Solution 3 would be a particular, documented extension to Dublin Core
that include a price element (and probably some others).  That would
solve the semantic interoperability problem (though, like all semantic
interoperability solutions, it will also require an out-of-band
understanding between the client and server).  No-one has suggested
this solution, and I am not suggesting it now: just mentioning it for
completeness.

Finally, there is the proposal that you posted, namely the ability to
extend Dublin Core records with elements from the Record Metadata
namespace.  I think this is a good and useful proposal, but it has
nothing to do with the particular problem we set out to solve.  And if
it is adopted, it should _not_ be called DCX as that name has been in
use for something different (solution 2) for the last few years.

So what do I recommend?

1. We could adopt your recent proposal, under a name like DC+RMD
   rather than DCX.  But maybe it's better not to do that since no-one
   has yet stated a need for it.

2. We could codify the DCX schema (Dublin Core records augmented by
   any non-DC elements), if only so that Theo can stop badgering the
   list about it every few months :-).  I don't see much actual harm
   in this, although I would not use it in my own applications.

3. The person who asked the original question should ignore all this
   and just use the ONIX schema, which was designed to do precisely
   what he is trying to do.

 _/|_	 ___________________________________________________________________
/o ) \/  Mike Taylor    <[log in to unmask]>    http://www.miketaylor.org.uk
)_v__/\  "I've thought ornithopods were by far the most painfully boring
	 dinosaurs for my entire life.  If I ever develop a powerful
	 sedative I'll name it Iguanodon" -- Matt Wedel.