Re: Central vs. Distributed Archives

From: Michael L. Nelson <mln_at_ILS.UNC.EDU>
Date: Thu, 2 Nov 2000 17:30:17 -0500

(note: I'm not sure this will get through all the aliases -- I don't think
this email addr is registered with the UPS list, for example)

On Thu, 2 Nov 2000, Steve Hitchcock wrote:

> NCSTRL was effectively the model for OAi. Greg Kuperberg suggests that
> NCSTRL has not been successful. It would be useful to have some meaningful
> measure of whether NCSTRL has been successful or not, and to hear the views
> of the NCSTRL developers (who are also involved in OAi). Maybe real
> evidence will yield clues to the ultimate destiny of OAi - central or
> distributed.

just a point of clarification: NCSTRL was not directly the model for OAI,
at least architecturally.

OAI has more in common with:

        - RePEc (
        - SODA (

and similar architectures.

A subset of the Dienst protocol gave us a starting ground for defining a
harvesting protocol, but even that has been relaxed to allow Dienst and
OAI to progress independently.

Most OAI service providers will probably assume a distributed storage
model, because it is certainly easier to build. But technically OAI is
agnostic with respect to centralized vs. distributed storage of data.
OAI focuses only on metadata.

Regarding centralized vs. distributed, I would submit CiteSeer

as an exemplary DL that seems to have resolved the tension between the two
models - providing both links to distributed copies and cached centralized



> The Harnad-Kuperberg dialogue has been fascinating but, to my mind, hasn't
> resolved the issue conclusively. It will be critical to understand what the
> user wants.
> Steve
> ------------------------------------------------------
> UPS mail list
> Mail submissions to
> To subscribe or unsubscribe visit

Michael L. Nelson
207 Manning Hall, School of Information and Library Science
University of North Carolina  
Chapel Hill, NC 27599         
+1 919 966 5042                         +1 919 962 8071 (f)
Received on Mon Jan 24 2000 - 19:17:43 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:45:56 GMT