Poul,

I can't quite follow what you are arguing for or against.

RDF is a language for specifying a metadata schema. To quote from the
specs it 'does not specify a vocabulary of descriptive elements such as
"author". Instead, it specifies the mechanisms needed to define such
elements, to define the classes of resources they may be used with, to
restrict possible combinations of classes and relationships, and to
detect violations of those restrictions. Thus, this document defines a
schema specification language. More succinctly, the RDF Schema mechanism
provides a basic type system for use in RDF models. It defines resources
and properties such as rdfs:Class and rdfs:subClassOf that are used in
specifying application-specific schemas.'

i.e. it performs much the same purpose as XML Schema (or even DTD), but
allows a more expressive description and hence stronger validating
parsers.

Having (if I recall) agreed that ZNG would assume that all records would
be in XML format and given that RDF also uses XML (indeed the current
RDF Schema spec.s are such that they are also XML Schemas if I
understand correctly) then in the case of ZNG, since the record
structure will be XML, ZNG can support RDF Schema specified record
schemas in the same way it handles XML Schema or DTD specified record
schemas.

In all cases (irrespective of whether the schema is specified as RDF
Schema, XML Schema or DTD, or for that matter as more than one of
these), we will still need to identify the record schema used in the
search and the response. Saying the default record schema is RDF Schema
is as meaningless as saying that the default record schema is XML Schema
or even is XML! You need to say which RDF Schema description is used (in
the same way you need to say which XML schema is used).

As such ZNG is agnostic about what mechanism is used to describe the
structure of the XML record providing that the structure can be named.

Matthew

> -----Original Message-----
> From: Poul Henrik Jørgensen [mailto:[log in to unmask]]
> Sent: 25 September 2001 20:34
> To: [log in to unmask]
> Subject: Re: metadata formats for ZNG
>
> Hi Ray and others,
>
> The main reason for supporting the RDF/XML format is, that RDF is
agnostic
> about element sets, thus avoiding getting ZNG entangled in feuds over
> which
> metadata element set to use.
>
> For metadata element sets, which does have an associated XML format,
it is
> natural to use this format. DCMI for example recommends the RDF/XML
format
> for the simple Dublin Core metadata element set.
>
> ZNG should only specify RDF/XML as "default" format for retrieval of
any
> metadata element set, which does not by itself have an official XML
format.
>
> Best regards,
> Poul Henrik
>
> -----Original Message-----
> From: Ray Denenberg [mailto:[log in to unmask]]
> Sent: 25. september 2001 19:22
> To: [log in to unmask]
> Subject: Re: metadata formats for ZNG
>
>
> Ok, Jan (and Matthew) , please disregard my reference to "replacing"
ONIX.
>
> We're proposing this new metadata set (called "yanks") to supplement
the
> two
> already proposed. Yanks is for bibliographic applications. It's
> development
> is
> being initiated in our office, not specifically for ZNG. It's a marc
> subset,
> with human-readable element names and corresponding marc tags
documented
> as
> annotation, in an XML schema. I expect to be able to provide a draft
by
> Thursday.
>
> Now in response to Poul Henrik, the thrust of our proposal is (1) the
> element
> set (2) the use of human-intelligible element names, and (3) the marc
> mapping.
> The format is not of primary importance, though a format is proposed
(an
> XML
> schema). I think we're confusing format and element set. ZNG has
(so
> far)
> listed two element sets, DC and onyx. Each comes with its own
"format".
> In
> general, ZNG hasn't adopted a format (other than to say that we'll use
XML)
> we're planning to use the default format for a given set. I don't see
> what
> additional value RDF gives us.
>
> With respect to the OAI marc schema, there may be good reason to add
that
> too,
> depending on what we decide is the relationship of ZNG and OAI.
However
> these
> are very different schemas. The OAI schema simply provides an XML-
> compatible
> way
> to transparently transfer marc data.
>
>
> --Ray
>
> Janifer Gatenby wrote:
>
> > I support Matthew's argument. When we see LC's new schema and know
its
> > purpose, I'm sure that we will want to include it. Nevertheless,
> including
> > ONIX was our message to the outside world that ZNG is suitable for
non
> > library / bibliographic systems. Therefore, at this stage, I would
not
> like
> > ONIX to disappear.
> >
> > Janifer
> >
>
> --
> Ray Denenberg
> Library of Congress
> [log in to unmask]
> 202-707-5795