SV: MarcXchange - proposal for a new standard for MARC-records

Mr. Andresen!
 
I've been looking at your schema and there are a few things which I'm
wondering about. I suppose more people may find it interesting, so perhaps
you may be so kind to write why you've decided so:
*) As far as I understand the idea basically is to have a singular
collection element which holds a number (perhaps zero) record elements. Your
schema makes it possible to have a singular record element without a
surrounding collection element. In this way there are two ways to represent
a singular record: With or without surrounding collection element. This make
it ambigous to parse by software. I think the structure should always be the
same no matter whether threre's only at most one record element or more
record elements. Don't you think that a record element should not be able to
exist without a surrounding collection element.
 
*) I may see that a collection element is nillable. I suppose that an empty
collection would just contain no record sub-elements so what's the point of
having the collection element to be nillale?
 
*) I may see that a record element is nillable. I suppose that if you do not
have anything to specify for a given record you would simply not include the
record element so what's the point of having the record element to be
nillable?
 
*) Wouldn't it be better to have the format attribute of the recordType
complextype to be an enumeration? In this way the schema could define what
could be found in concrete XML documents, which would make it much easier to
exchange MarcXchange documents between institutions. I suppose the set of
formats is well defined. 
 
*) Wouldn't it be better to have the type attributoe of the recordType
complextype to be an enumeration? In this way the schema could define what
could be found in concrete XML documents, which would make it much easier to
exchange MarcXchange documents between institutions. I suppose the set of
types is well defined.
 
*) Your schema allows a record element to avoid having a leader subelement.
Shouldn't a record element have exactly one leader element?
 
*) In your pdf file "n569.pdf". There's a sample XML document which is
claimed to be a valid MarcXchange document. Definitely it's a well-formed
XML document, but there should be a namespace specification for the elements
in order to make it valid.
 
*) I think you should enhance your namespace to include a version number,
e.g. to change it from "http://www.bs.dk/standards/MarcXchange" to "
http://www.bs.dk/standards/MarcXchange/1.0
<http://www.bs.dk/standards/MarcXchange/1.0> ". In this way it would be
possible to increase the version number as new versions may arise.
 
 
Best regards,
 
Henrik Dahl
Dantek

-----Oprindelig meddelelse-----
Fra: www-zig-request@w3.org [mailto:www-zig-request@w3.org]Pa vegne af Leif
Andresen
Sendt: 27. april 2005 15:39
Til: www-zig@w3.org
Emne: MarcXchange - proposal for a new standard for MARC-records



To ZIG-list 

In these months the ISO committee for Technical Interoperability -
Information and Documentation votes on a proposal for a New Work Item about
a XML schema to wrap MARC-records. For many years ISO2709 has worked well,
but today it is necessary to use XML in many cases. 

That's the reason for the proposal for MarcXchange. 

More information:  <http://www.bs.dk/marcxchange/>
http://www.bs.dk/marcxchange/ 


Best regards, 

Leif Andresen 
Chair of Danish Standard S24 Information and Documentation 

********************************************************* 
Leif Andresen    *   Email: lea@bs.dk 
Library Advisory Officer 
Danish National Library Authority 
Nyhavn 31 E, DK-1051 Copenhagen K 
Phone direct: +45 3373 3354 
Phone: +45 3373 3373 *  Telefax: +45 3373 3372 
*********************************************************