3 OVERVIEW OF PROGRESS IN LIBRARY DEVELOPMENT

3 OVERVIEW OF PROGRESS IN LIBRARY DEVELOPMENT

We rate our progress in terms of both research and development as very good, and beyond our initial expectations. The first ADL cycle of development, (10/01/94-02/28/95) ended successfully with the completion of our first testbed system, which was a stand-alone "rapid prototype" (RP) system (see the paper by Fischer, Frew, Larsgaard, and Smith, 1995). A key factor that has played an important role in the success of our development since the completion of the RP was our realization of the deep significance of the nature and rate of change of relevant Internet-related technologies.

In response to this realization, we formulated and adopted a strategy for the development of ADL that has had important implications for our progress during the past year. The main components of our strategy emphasize:

1  a focus upon access to the many classes of collection items, including non-traditional items, by geographical reference;

2  the development of the user interface (UI) and catalog components of the DL architecture;

3  accessibility to our catalog and collections via the Internet for a wide variety of users;

4  close interaction and interoperability with other DL activities by way of Internet-related technologies;

5  a process of incremental, evolutionary design and implementation of ADL that takes advantage of critical technological developments, and especially Internet-related technologies;

6  digitally-supportable extensions to traditional library functionality;

7  access to both the explicit and implicit information in DL collections.

Items 3)-5), in particular, have had a major effect on our development during the past year (03/01/95-02/29/96) and have led us, in particular, in to a second cycle of development involving the design and implementation of a second testbed based upon available World Wide Web (WWW) technology. We termed this testbed the "Web prototype" (WP).

The WP was brought to an initial completion by 11/01/95 and provided a superset of the RP's functionality via the WWW. Furthermore, our current research and development, and the research and development planned for the coming year (03/01/96-02/28/97) are predicated upon the available and soon-to-arrive extensions of these technologies.

Since the project began in 1994, we have completed two phases of testbed development and are currently involved in a third phase. These three phases involve respectively:

1.  the design and implementation of a stand-alone rapid prototype (RP) system (completed March 1995);
2.  the design and implementation of a WWW prototype (WP) system with restricted access (completed December 1995);
3.  the design and implementation of an extended WP system with public access (to be completed on or after July 1996).

The only significant developments relating to the RP system during the past twelve months has been the preparation, in conjunction with ESRI, of a CD containing a scaled-down version of the RP. Approximately 3000 copies of the CD were distributed in October 1995 to major producers and users of spatially-referenced information, together with a questionnaire that provides us with evaluative information about the system, and in particular about its functionality and user interface. This information is being used to help design the interface for various versions of the WP. It is unlikely that there will be any further developments relating to the RP.

Early versions of the WP were presented publicly in September (at the Alexandria Design Review) and November of 1995 (at the DLI meetings in Santa Barbara). These versions have provided the basis of our planning for testbed research development for the coming year. As may be noted from a comparison of our description of the WP testbed in the next section and a description of our plans for the next twelve months, much of our research and development involves modifying and extending the WP system.

Some of the modifications and extensions relate to preparing the testbed for public access by way of WWW in July 1996, and include major revisions to the nature of the interface; the use of Java and Java-like languages to support a much greater range of system functionality; the tuning of the DBMS to increase the performance of the catalog component with respect to spatial queries; the building of significant, accessible collections; and the integration of high-performance servers.

Other modifications and extensions relate to longer term issues of functionality and performance and include, for example, a major extension to our ability to access spatially-referenced materials by means of an "Alexandria Atlas"; a major effort to extend the expressiveness and functionality of the catalog component by means of a redesign of the catalog component, with respect to the nature and organization of the metadata that is used to support access; and major extensions of our image processing efforts.

Even longer term plans, which will receive significant attention during the period 1997-98, include the support of access to the information implicit in the collections of the library by means of information processing procedures applied to retrieved sets of library items.

Finally we note that we have begun a major effort to extend our activities to those required for constructing a fully operational digital library (DL).