[Ore-ac] News from OAI Object Reuse and Exchange

[Dr P. Murray-Rust]

On Apr 9 2007, Carl Lagoze wrote:

Thanks for this Carl,

(I hope the other memebers of the AC do not mind the reply being copied to 
all.)

>
> We attach a document that we have been working on over the last few 
> weeks. It outlines a conceptual framework for the OAI-ORE work. Although 
> the document will undergeven with o further revisions as a result of 
> ORE-TC feedback, Clifford Lynch urged us to share it with you, believing 
> that it will help you understand and evaluate our work.
>
> 
>
> We hope this document will help to keep you informed on the direction of 
> the OAI-ORE effort, and we look forward to your feedback.
>
I have read this once over and the document addresses important problems we 
have had. The lack of compound documents, packaging and bounding has been a 
major problem in our work, highlighted by the difficulty of ingesting 
entries into DSpace (and I suspect most other repository types). Moreover 
the compound entries we have put into DSpace are not always trivial to 
extract, often requiring (simple) code to be written.

In general the questions I would ask at this stage of any project are: * 
has this been theoretically addressed before (Hytime and topic maps are 
obvious examples). We spent a long time on XML-DEV addressing 
hyperdocuments. From what I can see this is a selected subset of complete 
hypermedia and I would endorse this. Hopefully it can be shown to be 
formally consistent. * are other groups working in this area (it's an 
obvious thing to do)? I have been very disappointed by the W3C's slowness 
to address compound documents - your document implies that there is no 
direct competition. Some formst such as PDF and MPEG allow containment of 
other documents but I don't know what support they are getting and whether 
they allow the same overlap between sets as ORE. * is it implementable? I 
am a great believer in "rough consensus and running code" and know that it 
is often more difficult to implement a design than appears at firest sight. 
If it *is* implementable, can the functionality be contained within a 
library, for otherwise everyone has to create their own implementations. Is 
it accessible to small projects who do not have the resources of large 
projects like DSpace, Fedora, etc.? * what is required from specialist 
domains? As an example, what do I have to provide to allow CML (Chemical 
Markup Language) components to be bound into an ORE object? XML schema? 
RDF? Ontologies?

and - as this is a milestone - how is the work likely to progress in the 
next six mnonths?

Looking foward to meeting in Phoenix.

P.

-- 
Peter Murray-Rust
Unilever Centre for Molecular Informatics
Chemistry Department, Cambridge University
Lensfield Road, CAMBRIDGE, CB2 1EW, UK
Tel: +44-1223-763069 Fax: +44 1223 763076