Exploring the potential of registries
  These community-built directories can help libraries, archives and museums manage and share data
By Eric Childress and Jeff Young
OCLC Programs and Research
Registry is a word that has multiple meanings and may be used as a label for many, often very different things—for example:
-
A wedding gift registry: a list of desired gifts a
to-be-married couple supplies to a retail store for
consultation by their wedding guests that helps later
gift buyers avoid duplicating already-purchased gifts.
-
The Domain Name Server (DNS) Registry: the
authoritative, master database of all domain names
registered in each Top Level Domain—this database
is the ultimate authority on IP (Internet Protocol) addresses
for all domains on the Internet.
 Common among most things that we refer to as “registries” are three attributes:
-
The registry is a list/database/collection that captures
specified information,
-
The registry carries the imprimatur of “authority”
(actual or presumed) and,
-
The registry supports specific uses, the default
use being consultation.
In practice, most registries also have a fourth attribute:
they are designed to enable the transfer of information
between multiple parties. In many cases the information
transfers are one-way: one party enters the data, and authorized
third parties may query the registry to retrieve
data. In other cases, the third parties may have privileges
to update information in the registry.
Libraries, museums and archives have long traditions of
building and maintaining registry-like tools including: the
accessions list (a record of titles/items added to a collection
with a number assigned by the collecting agency),
the shelflist (an up-to-date card catalog of library holdings,
typically arranged by call number and chiefly consulted by
library staff) and a registry of donors (a list of persons and
organizations donating items, money and/or other gifts).
While these hardcopy forms of recordkeeping have
largely been supplanted by electronic successors (and
many of these hardcopy registries have been given a
place of honor in the respective agencies’ archives), the
need to build, maintain and interact with recordkeeping
systems and registries has not diminished. Indeed, in a
world of networked, electronic information, this class of
tools is not needed less, but far more.
In OCLC Programs and Research, we’ve been exploring
various aspects of registries as a secondary,
enabling technology for a number of projects… but we
have also explored registry services as a central theme,
particularly through a now-archived project called WikiD.
 WikiD focused on developing and deploying a distributed
registry environment. To this purpose WikiD used
several standard technologies (primarily: wikis, OAI-PMH,
SRU and OpenURL) in novel combinations, and the result was the development
of an experimental repository
management system that was easy
to deploy, full-featured, customizable
and extensible. This system
could be implemented by a single
agency for its own resources, or
be deployed as middleware by any
agency to deliver new functionality
to existing repositories hosted by
other agencies, all with no action
by the hosting agency.
The experience and knowledge
gained with WikiD led to new insights
and new thinking about
registry and Web services design
that has influenced the design of
production systems, namely the
WorldCat Registry, and has now
coalesced into a formal design approach
we call the Registry Framework
Architecture (RFA).
Briefly, RFA is built on the
premise of reusable application
components that can be created
based on generic conceptual
models and minimal assumptions. Its implementation philosophy calls for:
-
using and remixing existing
standards such as OpenURL
(request model), SRU (search) to
deliver standards-based services
like OAI-PMH (harvest), and RSS (syndication);
-
simplifying development and
maintenance by using common
models (e.g., the OpenURL
model) and an enterprise-wide,
common identifier syntax; and
-
implementing services and
systems using open standards
and open-source technologies
whenever possible.
What results is a very efficient and
flexible software development environment
that allows OCLC to better
leverage its own software and services,
and more efficiently and swiftly
share and update registry information
within OCLC and with third parties.
RFA enables us to more easily build
new services and add features to existing
services by building upon existing
application components.
The initial implementation of this
approach has been the WorldCat
Registry, and we have plans to reimplement
several existing services to
comply with RFA. Going forward, new
services will also begin adopting this
approach.
OCLC registries now available or in development:
-
DLF/OCLC Registry of Digital Masters: a joint effort of the Digital
Library Federation and OCLC, it is a shared, central registry
of digitized and to-be-digitized materials, designed to support
preservation and mass digitization planning.
-
OCLC OpenURL Resolver Registry: a service for individuals
and information partners to maintain OpenURL resolver information
in a single location and use this information across services
in the information industry.
-
Registry of Copyright Evidence: a work in progress, this registry
will use data mined from WorldCat and received from other
sources to provide a tool to assist libraries and other agencies
in determining a work’s copyright status.
-
WorldCat Registry: a Web-based directory of libraries, archives
and museums that contains details about the physical and electronic
location of institutions, the relationships between them
and the services they provide.
 A WorldCat community | By the numbers
|
