Just to clarify, in terms of Preservation Repositories, or the preservation
service provider model, we don't envisage user deposit in these
repositories. Instead, a harvesting method is likely to be used. Of course
it's possible these services will create new versions, by migrating
formats, or updating metadata, but these would be under controlled, managed
conditions. That's the point of Preservation Repositories, a controlled
environment, based on an agreement between the IR and the preservation service.
Preserv Project Manager
IAM Group, School of Electronics and Computer Science
University of Southampton, SO17 1BJ, UK
Email: [log in to unmask]
Tel: +44 (0)23 8059 3256 Fax: +44 (0)23 8059 2865
At 15:38 31/03/2006, Chris Armstrong wrote:
>I think this is largely a matter of semantics - the nouns 'archive' and
>'repository' both imply long term, secure, maintained storage of data;
>the term 'open access' means free availability.
>If the community is going to continue using those adjectives with one or
>other noun, then confusion is bound to result as to the functionality
>required. Personally, I see no problem with both functions in a single
>repository ... and considerable advantages.
>My problem in the idea that there should be a split, with open access
>facilitated on the one hand and preservation/archives on the other,
>comes - once again - from version control. We already have a situation
>where a researcher may deposit a copy of an article in an institutional
>archive as well as in one or more subject archives (such as E-LIS);
>while all co-authors may do the same. Now we double the trouble with
>copies in OA *and* in Preservation Repositories. LOCKSS (Lots of Copies
>Keeps Stuff Safe) gone mad.
>Peter Suber wrote on this almost exactly 2 years ago: The many-copy
>problem and the many-copy solution
>Despite the title, he does not really have a solution but he does
>highlight the problems. The JISC is also concerned and has recently
>commissioned research into the control of versions in institutional
>The issue - knowing whether you are reading the most recent and most
>accurate version - remains even when dating or metadata is incorporated.
>One answer is bibliographical control, where a single institution - The
>British Library, for example - maintains the archive, and legislation is
>in place to begin this process. It seems to me that what we need is a
>single institutional archive with open access and self-deposit built
>into the system.
>Information Automation Limited
>t. (+44) 1974 251302
>e. [log in to unmask]