The search and retrieval of information in online databases is not a new concept. Libraries, government agencies, and similar organizations have been organizing metadata into searchable, distributed catalogs for decades. Many of these groups employ the Z39.50 standard communications protocol for information discovery when creating their systems and search mechanisms. This standard is widely used by library systems and research institutions, including the Library of Congress. This standard is also used by the FGDC Geospatial Data Clearinghouse, which is part of the National Spatial Data Infrastructure. The goal of the NSDI is to facilitate the sharing and use of spatial data both inside and outside the U.S. government.
With the ArcIMS Z39.50 Connector, your ArcIMS Metadata Service can participate in an existing library system or the FGDC Clearinghouse. Documents available from your Metadata Service can be searched by any Z39.50 client application.
Adding the Z39.50 Connector to your Metadata Service can make your metadata available to a wider audience with little effort. Many university, college, and public library systems and research institutions use Z39.50 clients to let people search and browse their catalogs. This connector lets you seamlessly integrate metadata describing your GIS resources into these existing search and discovery systems, and best of all, these users won’t need to learn how to use a new application.
Once you’ve made your metadata available to Z39.50 clients you can:
The Z39.50 protocol—formally named Information Retrieval (Z39.50); Application Service Definition and Protocol Specification, ANSI/NISO Z39.50-1995—specifies the rules of communication between a search client and an online database. The protocol was initially produced by the National Information Standards Organization (NISO), the standards development group serving the fields of publishing, libraries, and information services. The protocol’s name came about because NISO was once the Z39 committee of the American National Standards Institute (ANSI) and it was the fiftieth standard developed by that committee.
The Z39.50 protocol formalizes rules for querying, retrieving records, and asking for those records in a particular format. It doesn’t specify implementation details for clients and servers, but rather the language that clients and servers are expected to speak.
Today, the United States Library of Congress hosts the Z39.50 Maintenance Agency, and a group called the Z39.50 Implementors Group (ZIG) continues the standard’s development. The specification for the protocol is available at
http://lcweb.loc.gov/z3950/agency/ along with other Z39.50 resources.
Z39.50, just like ArcIMS, follows a client/server model—one computer requests something from another computer, which then processes the request and returns a response. This is a simple proposal when both client and server use the same language, terminology, and database organization, but not that easy in real-world scenarios.
Z39.50 establishes a method of communication that supports information retrieval in a distributed environment. As long as a server understands Z39.50 requests, a client can request information from its database without prior knowledge of its organizational structure. Because the standard doesn’t specify implementation details, the server can translate the query into whatever form is appropriate for its database.
Servers generally implement a profile of the Z39.50 protocol. A profile is a subset of the functions available in the Z39.50 specification that support a particular type of application, function, or community. Libraries generally use the bibliographic, BIB-1, profile. Online GIS repositories generally follow the GEO Profile. Established by the FGDC, the GEO Profile extends the bibliographic profile to support searching for geospatial metadata. The GEO Profile is the language used to search servers participating in the NSDI Clearinghouse.
The ArcIMS Z39.50 Connector operates as the server in the Z39.50 framework; it supports the GEO Profile. Requests sent by Z39.50 clients are translated from the GEO Profile into ArcXML, the language used by ArcIMS. The ArcXML request is then sent to a Metadata Service and a response is generated. The Z39.50 Connector formats the response according to the client’s request and sends it back to the Z39.50 client. This process is discussed in more detail in ‘Customizing the connector’ later in this chapter. For more information on ArcXML, refer to ArcXML Programmer’s Reference Guide.
If you don’t have an existing Z39.50 client, a revised version of the MCNC Center for Networked Information Discovery and Retrieval (CNIDR®) ISite Z39.50 software suite is freely available through the FGDC Clearinghouse and can be used to test your Z39.50-enabled site. This version of ISite, modified to support the GEO Profile, can be downloaded from:
http://www.fgdc.gov/clearinghouse/tutorials/getisite.html.
The ISite software suite contains clients as well as utilities to index and serve metadata. However, while you can effectively use the clients included with ISite to test your Z39.50 configuration, its server software, like most Z39.50 servers, utilizes port 210. The ArcIMS Z39.50 Connector functions as the Z39.50 server and, by default, also uses port 210. Disable ISite’s Z39.50 service or use separate ports to avoid the conflict caused by using both applications.
The Z39.50 protocol’s specification, related tutorials, and other information are available from the Z39.50 Maintenance Agency site, hosted by the United States Library of Congress at:
http://lcweb.loc.gov/z3950/agency/.
In addition, the ESRI implementation of the Z39.50 specification is based on the GEO Profile. For information, refer to:
http://www.blueangeltech.com/standards/GeoProfile/geo22.htm.
A number of Z39.50 Web gateways or clearinghouses are also available, including:
FGDC NSDI Clearinghouse—http://www.fgdc.gov/clearinghouse
Library of Congress gateway—http://lcweb.loc.gov/z3950/