Andy Powell wrote (a while ago):
> What
> I find much harder to live with is the fact that currently,
> any project or
> standards initiative that chooses to use XML rather than
> RDF/XML gets no
> guidance or help from DCMI - nuffin', zero, ziltch - other
> than a message
> that they've made the wrong design decision and that they
> should go back
> and use RDF. This is *not* acceptable IMHO.
For example, these days I'm hanging out with the geographic
information crowd, who are working around 20 standards through
ISO in Technical Committee 211. One of these is ISO 19115
(currently a Draft International Standard - i.e. only typos
can be changed now) covers Geographic Metadata.
It has >200 elements which, when scoping is taking into
account, can lead to metadata "records" with 100's of elements.
(N.B. it is very much tuned for "collection-level" metadata,
as opposed to "item-level".)
After MUCH debate they settled on UML for modelling, and
a serialisation of this in XML using a specified method for
mapping the metamodel. There is a whole standard devoted
to the mapping (ISO 19118) - though it is only in
"committee draft" form and is currently stymied due a
difference of opinion with the OpenGIS Consortium,
who are backing an XML serialisation that is actually
patterned much more on RDF! But even that does not
follow any of the RDF syntaxes that have surfaced from W3C.
> Note, I am not arguing for or against RDF here. I'm saying
> that you (the
> W3C) have not yet convinced people to use RDF. Until you do,
> DCMI has to
> recognise that there are a significant number of people
> implementing DC in
> 'plain' XML. It's no good saying to people who have decided
> not to use RDF
> that they 'should use RDF' because they've already decided
> not to use RDF!
> And very often that decision is made on religious grounds rather than
> technical grounds. This isn't right or ideal... but it is a
> fact of life.
Indeed: no way will you win a fight with ISO by insisting
they use some particular RDF syntax. There are some very
very hard heads involved, who look on XSD with some distaste
already and will leave the room if you bring some other RDF
in with you.
And meanwhile, OpenGIS Consortium are happy to use anyone's
Metadata schema - providing it can be expressed in XSD.
Did I say all this already some time?
Simon
|