Begin forwarded message:

From: "Daniel R. Rehak" <[log in to unmask]>
Date: 5 April 2005 15:22:22 GMT+02:00
To: <[log in to unmask]>, <[log in to unmask]>
Subject: RE: CORDRA
Reply-To: <[log in to unmask]>

Thanks to Lorna for forwarding this to me.
I'm responding directly to Andy and Lorna, but please forward to the list
since I don't think I can currently post.

The document that Andy cites has not been updated in a while; it is a
general overview.  The additional details that Andy asks about are not
formalized yet.  Our work has been to build and test approaches before
making formal requirements.  But also, as Lorna indicates, the general
approach is that CORDRA provides an overall model, and that each community
and each registry will define which specs they will mandate.

We currently have two registry projects underway.  The ADL Registry is in
late alpha testing.  We exploring issues around QoS (multiple mirrors for
storage, resolution, etc. in multiple locations), search robustness (testing
with thousands of metadata records), internationalization issues, edge
tests, etc.  For example, the ADL-R uses a LOM profile, and has a defined
protocol for talking to the registry.  Repositories have their own interface
protocols, e.g., some use OAI, some use private ones.  Some repositories use
LOM, some use DC.  There is then an intermediate layer where the repository
protocol is mapped to the registry protocol and the repository metadata is
mapped for the registry instance.  Again, some repositories use REST, some
use SOAP/WSDL for communications.  And there is yet an addition decision of
where the intermediate layer executes -- at the repository, at the registry,
or somewhere in between.  A lot of the work is to make sure that these
combinations work, so we can make recommendations on how to let the
different communities and components use their own approaches and still get
interoperability.  All that said, we do have preferred approaches, e.g.,
OAI, LOM and DC, SRW/SQI, SOAP Doc, Wrapped Literal.

The second registry project is just starting.  It will explore many more
similar issues, like how to support different query interfaces, how to
express policies, provenance, versioning, etc.

If anyone wants the most current overview, and not read the big document,
there is a paper for WWW2005 on the CORDRA site
http://cordra.net/cordra/information/publications/2005/www2005/cordrawww2005
.pdf

I had hoped to get something out for the DC conference, but missed the
deadline with other work.

The "documents" part of the site contains the public technical documents as
we produce them.

We are also formalizing plans for how to let other registry projects join
the overall effort since we have several others who would like to build
their own registries.  This community will help further define CORDRA.

        - Dan



-----Original Message-----
From: Lorna M. Campbell [mailto:[log in to unmask]]
Sent: Tuesday, April 05, 2005 04:00 AM
To: [log in to unmask]
Cc: [log in to unmask]
Subject: Re: CORDRA

Hi Andy,

As I understand it CORDRA is an ongoing work in process but
I'm really not the best person to comment on which parts of
the model are specified in what kind of technical detail so
I've forwarded your mail to Dan Rehak who should be able to
answer your query.

However I believe that CORDRA is a relatively open reference
model.  For example the ADL pilot implementation uses Handles
but the CORDRA model itself does not specify which identifier
solution should be used.

All the best
Lorna



On 4 Apr 2005, at 23:56, Andy Powell wrote:

I'm interested in CORDRA, in particular in finding out where
the technical specification of CORDRA has got to.  Looking at

CORDRA: Technical Introduction and Overview
http://www.lsal.cmu.edu/lsal/expertise/projects/cordra/intro/i
ntro-v1p00.html

I don't see much detail.  For example, in the section on
'Content Repositories' the document says

 "A content repository provides a standard interface to
retrieve an  identified content object. Content repository
registration will include  the specification of the content
exchange methods."

but doesn't say anything more about what the standard
interface is (thopugh it sounds a little bit like an HTTP GET
or possibly an OAI-PMH GetRecord request with the requested
format indicating some kind of complex object such as a METS
package or similar?).  Is this level of detail not defined
yet... or am I looking in the wrong place?

Thanks,

Andy
--
Distributed Systems, UKOLN, University of Bath, Bath, BA2 7AY, UK
http://www.ukoln.ac.uk/ukoln/staff/a.powell/      +44 1225 383933
Resource Discovery Network http://www.rdn.ac.uk/







--
Lorna M. Campbell
Assistant Director
Centre for Educational Technology Interoperability Standards (CETIS)
Centre for Academic Practice, University of Strathclyde
+44 (0)141 548 3072
http://www.cetis.ac.uk/