Stephane Sechaud wrote:

> Isn't it just a case of creating customized metadata to suit 
> the needs of these learning resource repositories? DSpace 
> supports custom metadata already does it not? Wouldn't it be 
> easier to create this metadata and then build services on top 
> of it rather than create a whole new repository software from scratch?

IMO, wearing my database/web apps developer hat, no. If I were designing
a database, in the sense of analysing requirements then producing
entity-relationship diagrams, I'd produce different designs for
'binaries' and 'scholarly materials' repositories. To take a very simple
example, in a repository for e-learning materials (a small-scale example
of which I've recently produced), I'd have tables for filetype and
mimetype, and have relations to these from the main item record so that
end-users could search for materials of particular type(s). If I were
producing something even more specific, such as an image management
system, I'd have image-specific tables and relations and cater for
standard schemas such as EXIF. Now, in an abstract sense this is just
"customized metadata", but the practical implementation requires, IMO,
specific database designs for particular repository material types. 

Apologies to all if I'm getting the wrong end of the stick, and seeing
the trees rather than the wood. I'm primarily a technical developer and
tend to see things at quite a low level. I also have no experience of
DSpace, other than evaluating it briefly a while back, so can't say how
well, or not, it could store, catalogue and deliver e-learning
materials.

Cheers

Fred Riley
Learning Technologist
School of Nursing, Midwifery and Physiotherapy, University of Nottingham
Vcard: http://www.nottingham.ac.uk/nursing/sonet/about/fr_uon.vcf  

> 
> Stephane
> 
> > --------------------------------------------------
> > From: "Fred Riley" <[log in to unmask]>
> > Sent: Monday, January 26, 2009 5:24 PM
> > To: <[log in to unmask]>
> > Subject: Re: List for learning object repositories?
> >
> >> Leslie Carr wrote:
> >>
> >>> I have looked at the product definition for intraLibrary, 
> and it is 
> >>> very difficult to distinguish it from a generic repository. 
> >>> Obviously the devil is in the details, but IN THEORY, I don't see 
> >>> that there is a fundamental distinction between them.
> >>
> >> <shrug> Theory's one thing, practice is another. At a high 
> level of 
> >> abstraction I'm sure you're right, but on a practical level I 
> >> wouldn't use DSpace, or any other repository designed specifically 
> >> for scholarly materials, to store, catalogue and serve e-learning 
> >> resources, and I wouldn't use Intralibrary, for example, for 
> >> scholarly works. And I'm afraid that this techie operates 
> very much 
> >> at the mundane, practical level, which is why I'd like to 
> see a forum 
> >> where practical learning materials repository issues can 
> be thrashed 
> >> out. Which forum, as Sarah has pointed out, does already exist in 
> >> cetis-metadata, the name of which made me forget that learning 
> >> materials repositories are very much in CETIS' remit.
> >>
> >>> > Er, precisely because "a repository is a database and some
> >>> storage and
> >>> > a heap of services". To store, catalogue and serve learning
> >>> resources
> >>> > the database needs to be designed rather differently 
> from that of 
> >>> > a scholarly materials repository, and the services on 
> top need to 
> >>> > be tailored to learning resources.
> >>> Different schema, different workflows, different services.
> >>> That's just customisation!
> >>
> >> I'm a database developer, amongst other things, and I wouldn't 
> >> describe a completely different database schema as "just 
> >> customisation", and for sure the database design I'd come 
> up with for 
> >> publications would be very different from that for e-learning 
> >> materials. It's a different database for a different 
> purpose. Again, 
> >> at a high level of abstraction all repositories are but 
> instances of 
> >> the same ideal meta-repository, an 'abstract class' 
> perhaps, but then 
> >> at a high level of abstraction we might be just programs 
> running in a 
> >> universe-wide quantum computer ;-)
> >>
> >> Cheers
> >>
> >> Fred Riley
> >> Learning Technologist
> >> School of Nursing, Midwifery and Physiotherapy, University of 
> >> Nottingham
> >> Vcard: http://www.nottingham.ac.uk/nursing/sonet/about/fr_uon.vcf
> >>
> 

This message has been checked for viruses but the contents of an attachment
may still contain software viruses, which could damage your computer system:
you are advised to perform your own checks. Email communications with the
University of Nottingham may be monitored as permitted by UK legislation.