Interestingly, as I understand it the new biblios.net cooperative
metadatastore from LibLime takes this approach. I don't know if it uses
XSLT or not, but if I understand correctly, the underlying data store
supports having multiple schema-representations of the same resource,
and relating those different schemas (say MARC, DC, ONIX, some future
RDA-inspired schema, etc--although I don't think it supports all of
these YET, the architecture does). An individual schema representation
_can_ be actually individually uploaded, and the store will keep track
of it's relation to other representations of the same resource. But if a
representation in a particular desired format hasn't been uploaded
individually---biblios.net has the capability of automatically
translating an existing representation into the desired representation.
I find this approach very exciting.
Jonathan
Weinheimer Jim wrote:
And as an extra wrinkle when considering these matters, the original file may
be in XML and the different formats are merely generated from this one file
through the use of different XSL Transformations. So, a single file in XML
format can create *on the fly* a pdf, html, MSWord, and theoretically, anything
else you would like.
As a very simple example of how this works in action, see:
http://www.w3schools.com/xsl/xsl_transformation.asp, which takes an xml file
and transforms it using your browser into an html document. But, correctly
coded, any XML file document could even be turned into a MARC record!
I think this will be happening more and more often since it is such a waste of time and
resources to literally store multiple formats, when they can simply be generated by
modern browsers. And, these abilities can become extremely powerful. This is an example
of information resources that are genuinely new; these resources don't really
"exist" except as viewed on your machine at a specific moment. Change the XSLT
and what you see on the screen changes completely, but the XML file does not.
It's interesting from a theoretical point of view since the idea of "manifestation"
disintegrates completely, while the idea of "expression" becomes tangible in the XML
file. But how to deal with it in our everyday workflow remains to be solved.
Jim Weinheimer
Jakob asked:
>For instance how to you catalog a a web page or a PDF file deposited at
>a repository? And what do you do if the PDF is also availabe at a second
>server and there is a HTML version of the same document at the third?
RDA Appendix M has examples:
http://www.rdaonline.org/constituencyreview/Phase1AppM_11_10_08.pdf
There are examples for an audio book, audio music, two books, a
serial, a streaming video, a video recording, and a website. There is
no example of an electronic document; neither "electronic text" nor
"electronic document" are in RDA's list of possibilites at RDA 3.3 p.
12, only the very general "online resource", which is more like a GMD
than an SMD. Electronic documents make up the bulk of our work these
days, and often exist in the variety you mention. RDA is distant from
the bibliographic world we experience, and the needs expressed by our
clients.
Regardless of how RDA would have you do it, I can tell you how we do
it, and will probably continue to do it.
There seem to be four possibilties: (1) a record for each format
(driving reference and patrons mad); (2) a record for one format with
the others in 530's; (3) repeating collations and 856's, one for each
format (300 is repeating because of the British way of doing kits);
and (4) a generalized 300* with repeating 856's, or an 856 which takes
you to an introductory page which has links to the various formats.
*300 $a1 electronic text (x, 100 p. : graphs) :$bdigital file.
We use the last of the four methods mentioned above. Offlist I will
send you the cataloguer instructions for one such project.
&nbs
p; __ __ J.
McRee (Mac) Elrod ([EMAIL PROTECTED])
&nbs
p;{__ | / Special
Libraries Cataloguing HTTP://www.slc.bc.ca/
___} |__ \__________________________________________________________
--
Jonathan Rochkind
Digital Services Software Engineer
The Sheridan Libraries
Johns Hopkins University
410.516.8886
rochkind (at) jhu.edu
