ANSI/NISO Z39.50-1992                          ISSN: 1041-5653
(Revision of ANSI/NISO Z39.50-1988)

Application Service Definition and Protocol Specification for Open Systems Interconnection

Foreward

(This foreword is not part of American National Standard for 
Information Retrieval Application Service Definition and Protocol 
Specification for Open Systems Interconnection, ANSI/NISO 
Z39.50-1992.  It is included for information only.)

Version 1 was prepared by Committee D, "Computer-to-Computer Protocols," of NISO, the National Information Standards Organization. Committee D was organized to develop OSI application layer protocols for library applications, including an information retrieval protocol. The committee was disbanded after Z39.50 version 1 was approved. In 1989, the Z39.50 Maintenance Agency was formed, administered at the Library of Congress. Version 2 has been prepared by the Maintenance Agency to incorporate enhancements described in detail below.

In 1990 the Z39.50 Implementors Group (ZIG) was established. Members include manufacturers, vendors, consultants, information providers, and universities, who wish to access or provide access to various types of information, including bibliographic, full text, financial, public utility, chemical, and news. Thus, although the protocol was originally proposed (in 1984) for use with bibliographic applications, interest in Z39.50 as an information retrieval protocol is no longer limited to use with bibliographic information.

Z39.50 version 2 represents a consensus of the ZIG, which has in effect acted in an advisory capacity to the Maintenance Agency in the effort to develop version 2. ZIG Membership has been open to all interested parties.

Information Retrieval Protocol

The protocol specifies formats and procedures governing the exchange of messages between a requesting computer (the "origin") and a responding computer (the "target") to enable the origin to (1) request that the target search a database and identify records that meet specified criteria, and (2) request transmission of, and receive, some or all of the identified records.

The origin may initiate requests on behalf of a user who wishes to search a database located on the target system. The protocol addresses communication between corresponding information retrieval applications on the origin and target systems; it does not address interaction between the origin computer and user.

Basic Operations

The Information Retrieval protocol provides the following basic capabilities: The origin may submit a Search request which includes a query, and parameters that determine whether or not records resulting from the search are to be returned as part of the response. The target responds with a count of records identified and possibly some or all of the records. The origin may then submit a Present request, requesting transmission of selected records. The origin assumes that records selected by the search request form an ordered set (the "result set") which may be referenced by sequential position within the set. Record order is determined by the target. The origin may request (for example) records one through four, and follow with a request for records four through eight, and then records two through three, etc. The origin may submit as many such Present requests as desired, and may then submit another Search request.

Optional capabilities include the following:

• The origin may specify element set names indicating data elements that should be transmitted in cases where the origin does not wish to receive complete database records. For example, the origin might specify "If 5 or fewer records are identified, transmit `full' records; if more than 5 records are found, transmit `brief' records." (In this example, the meanings of `full' and `brief' are target specific.)

• The origin may specify a preferred-record-syntax for response records. For example, it might identify a bibliographic format such as USMARC.

• The origin may name a result set for subsequent reference.

• The origin may delete a named result set.

• The target may impose access control restrictions on the origin, by demanding authentication before processing a request.

• The target may provide resource control by sending an unsolicited or solicited status report. This might occur during processing of an operation, in which case the target may suspend processing and allow the origin to decide whether the operation should continue.

This standard mandates support of the "Type-1" query as a common format for the intersystem transmission of a search query. The "Type-1" query, fully specified in this standard, expresses a search query by individual search terms (or phrases) with a set of attributes for each. Attributes may specify, for example, type of term (subject, name, etc.), whether it is truncated, and its structure. The target is responsible for mapping these attributes to the logical design of the database.

Several terms may be combined in a Type-1 query, linked by Boolean operators. Terms and operators are expressed in Reverse Polish Notation.

Attribute Sets

The attributes associated with a search term belong to a particular attribute set, whose definition assigns integer values to various attributes. The definition is assigned a unique and globally recognized "attribute-set-id," an OSI Object Identifier, which is included within the query.

The process of assigning an OSI object identifier is called "registration." Appendix C defines and registers one attribute-set-id, "bib-1," which specifies various attributes that are useful for bibliographic queries. There is a mechanism to register additional attribute sets. Just as the bib-1 attribute set was developed by the bibliographic community, it is intended that attribute sets will be developed and registered as needed by other communities (for example, a chemical attribute set by chemists).

Response Records

The protocol provides for the transfer of database records identified by a Search or Present request. The standard distinguishes two types of records, database and diagnostic records, which can occur in response messages from the target.

Appendix E registers object identifiers for various MARC formats, including USMARC, UKMARC, Norway MARC and CANMARC; these object identifiers accompany database records returned by the target. There is a provision for registration of additional record formats, including non-bibliographic formats.

Diagnostic records are similarly accompanied by an object identifier that identifies their format. Appendix D defines and registers one diagnostic record format, "bib-1," which includes various diagnostic codes that are useful for bibliographic applications. Additional diagnostic record formats may be registered.

Changes in Version 2

Z39.50 version 2 incorporates major enhancements to version 1, in two categories:

• changes necessary for alignment with ISO 10162/10163, the Search and Retrieve (SR) Service Definition and Protocol Specification; and

• incorporation of features deemed necessary by implementors, to provide sufficient functionality so that implementation can be economically justified.

When SR was approved in 1991, incompatibilities between Z39.50 version 1 and SR remained:

• SR Search and Present requests include parameters (related to record composition and syntax) not included in Z39.50 version 1. Specifically: (1) the search request in SR has both small-set- and medium-set-element-set-name parameters. Z39.50 version 1 has only a single, or global, element-set-name parameter. (2) SR includes the parameter Preferred-record-syntax, and Z39.50 version 1 does not.

• The SR Delete service allows the origin to specify a list of result sets to be deleted. Z39.50 version 1 allows deletion of a single specified result set only.

• SR uses ASN.1 ("Abstract Syntax Notation one") to define the abstract syntax of its protocol data units, and specifies ASN.1 Basic Encoding Rules for transfer syntax. Z39.50 Version 1, finalized before ASN.1 was stable, developed a non-standard abstract syntax notation and corresponding transfer syntax, which in version 2 have been abandoned in favor of ASN.1.

• A major limitation of Z39.50 version 1 is the lack of flexibility to reference objects, including attribute-sets, diagnostic-sets, and record formats, which SR references through OSI object identifiers.

The above differences are resolved in Z39.50 version 2. There will still be differences with SR, however. Z39.50 includes two services not yet in SR: access control and resource control. Some U.S. information providers cannot use Z39.50 if it does not include an access control service. Both services have been carefully incorporated so that an SR and Z39.50 version 2 origin/target pair may interwork transparently (i.e. the SR implementation would not be aware that its partner is a Z39.50 rather than SR implementation).

In general, implementors who plan to build Z39.50-specific features into their implementations, for use when interworking with a Z39.50 partner, would suppress these features when interworking with an SR partner. Thus, lack of these features in SR will not inhibit interoperability, although it might limit potential functionality. For example, information that an SR origin can receive from a Z39.50 target might be limited because the origin cannot respond to an access control request. (The same would be true of a Z39.50 origin which does not support access control--it is an optional service). The limitation would not owe to an incompatibility between origin and target implementations, but rather, because the target requires authentication and the origin cannot provide it.

Version 2 includes changes to the Resource Control facility. The origin will be able to request a resource control action or cancel an operation. These changes have been expressed as requirements by the ZIG.

Version 2 specifies five query types. Types 0,1, and 2 are identical to the three query types specified in SR. The "Type-0" query is designated "private," allowing two systems to use a private, mutually agreed upon query format. "Type-1" is described above. "Type-2" is as specified by ISO 8777. Z39.50 specifies two additional query types: "Type-100" is as specified by ANSI Z39.58, and "type-101" is for proximity searching. The latter is specified in Appendix G. Proximity searching is a feature that implementors recommended be incorporated into this version.

Object Identifiers

Z39.50 registers all information objects that are registered in SR: application context, abstract syntax, attribute set, diagnostic set, and record syntax definitions. In addition, Z39.50 attribute set and diagnostic set definitions contains additional attribute and diagnostic values that are "U.S. specific."

Z39.50 registers two additional object identifier classes: resource report format (for resource control, which is not yet in SR), and transfer syntax. The latter accommodates non-bibliographic databases. Z39.50 also provides a mechanism for registration of experimental and implementation-specific objects. The Z39.50 Maintenance Agency will act as registration authority for Z39.50 objects.

Future Services

The generality of the protocol is intended to allow accommodation of new services as they are required. Some services being considered are:

• Explain: to allow an origin to obtain details of the target implementation, including databases available for searching, data elements in records of a specified database, element sets supported, elements within in a specified element set, attribute sets supported, attributes supported within a supported attribute set, diagnostic sets, and record syntaxes.

• Define Element Set: to allow an origin to define and name a set of elements to compose a retrieval record.

• Scan: to allow an origin to obtain a list of access point values from a database index (subject terms, names, titles, etc.), preceding and following a specified access value.

• Object Access: To allow an origin to perform operations including create, delete, sort, export, activate, and permit, on various objects, including result sets and queries.

• Segmentation: to allow a target to return multiple responses to a single present request.

Suggestions for improving this standard are welcome. They should be sent to the National Information Standards Organization, P. O. Box 1056, Bethesda, MD 20827, (301) 975-2814.

This standard was processed and approved for submittal to ANSI by the National Information Standards Organization. NISO approval of this standard does not necessarily imply that all Voting Members voted for its approval. At the time it approved this standard, NISO had the following members:

   American Association of Law Libraries
     Gary J. Bravy

   American Chemical Society
     Robert S. Tannehill, Jr.
     Leon R. Blauvelt (Alt)

   American Library Association
     Myron Chace
     Sally McCallum (Alt)

   American Pyscological Association
     Lois Granick
     Maurine F. Jackson (Alt)

   American Society for Information Science
     Clifford Lynch

   American Society of Indexers
     Jessica Milstead
     Patricia S. Kuhr (Alt)

   American Theological Library Association
     Raymond Vandegrift

   Apple Computer, Inc.
     Karen Higgenbottom

   Art Libraries Society of North America
     Patricia J. Barnett
     Pamela J. Parry (Alt)

   Association of American Publishers (AUP)
     Sandra K. Paul
     Thomas D. McKee (Alt)

   Association of American University Presses (AAUP)
     Mary Lou Menches

   Association of Information and Dissemination Centers ASIDIC
     Bruce H. Kiesel

   Association of Information and Image Management AIIM
     Mary Courtot

   Association of Jewish Libraries
     Bella Hass Weinberg
     Pearl Berger (Alt)

   Association for Recorded Sound Collections (ARSC)
     Donald McCormick
     Barbara Sawka (Alt)

   Association of Research Libraries (ARL)
     Duane E. Webster

   AT&T Bell Labs
     George E. Grant
     M. E. Brennan (Alt)

   Baker & Taylor Books
     Christian K. Larew
     Stephanie Lanzalotto (Alt)

   Book Manufacturers' Institute
     Douglas Horner

   Catholic Library Association
     Michael B. Finnerty

   CLSI, Inc.
     Bob Walton
     Susan Stearnes (Alt)

   Colorado Alliance of Research Libraries (CARL)
     Ward Shaw

   Council of National Library of Information Associations
     Thomas J. Kemp

   Data Research Associates, Inc.
     Michael J. Mellinger
     James Michael (Alt)

   Dynix
     Rick Wilson

   EBSCONET, Inc.
     Sharon Cline McKay
     Mary Beth Vanderpoorten (Alt)

   Engineering Information, Inc.
     Eric Johnson
     Mary Berger (Alt)

   The Faxon Co., Inc.
     Fritz Schwartz
     Joe Santosuosso (Alt)

   Gaylord Information Systems
     Robert Riley
     Bradley McLean (Alt)

   IBM Corporation
     Peggy Federhart

   Indiana Cooperative Library Services Authority (INCOLSA)
     Barbara Evans Markuson
     Janice Cox (Alt)

   Library Binding Institute
     Sally Grauer

   Library of Congress
     Henriette D. Avram
     Sally H. McCallum (Alt)

   Mead Data Central
     Peter Ryall
     Dave Withers (Alt)

   Medical Library Association
     Rick B. Forsman
     Raymond A. Palmer (Alt)

   MINITEX
     Anita Anker Branin
     William DeJohn (Alt)

   Music Library Association
     Lenore Coral
     Geraldine Ostrove (Alt)

   National Agricultural Library
     Joseph H. Howard
     Gary K. McCone (Alt)

   National Archives and Records Administration
     Alan Calmes

   National Institute of Standards and Technology 
   (NIST), Office of of Information Services,
   Research Information Center
     Marietta Nelson

   National Library of Medicine
     Lois Ann Colaianni

   OCLC, Inc.
     Kate Nevins
     Don Muccino (Alt)

   OHIONET
     Joel Kent
     Greg Pronevitz (Alt)

   Optical Publishing Association
     John Nairn
     R. Bowers (Alt)

   PALINET
     James E. Rush

   Pittsburgh Regional Library Center (PRLC)
     H. E. Broadbent III

   Readmore Academic Services
     Sandra Gurshman
     Dan Tonkery (Alt)

   The Research Libraries Group, Inc.  RLG
     Wayne Davison
     Kathy Bales

   Society of American Archivists
     to be appointed

   Special Libraries Association  SLA
     Audrey Grosch

   SUNY/OCLC Network
     Glyn T. Evans
     David Forsythe (Alt)

   UMI
     Don Willis

   Unisys Corporation
     Michel Ridgeway
     James Thomas (Alt)

   U.S. Department of Commerce 
   Printing & Publishing Division
     William S. Lofquist

   U.S. Department of Defense, Army
   Library Management Office
     Dorothy Fisk McDysan
     Diane Zehnpfennig (Alt)

   U.S. Dept. of Defense Defense Technical 
   Information Center (DTIC)
     Claire Tozier
     Gretchen Schlag (Alt)

   U.S. Dept. of Energy, Office of Scientific 
   & Technical Information
     Mary Hall
     Nancy Hardin (Alt)

   U.S. ISBN Maintenance Agency
     Emery Koltay

   U. S. National Commission on Libraries and 
   Information Science (NCLIS)
     Peter Young
     Sandra N. Milevski (Alt)

   VTLS, Inc.
     Vinod Chachra

   H. W. Wilson Company
     George I. Lewicky
     Ann Case (Alt)

At the time NISO approved this standard, the following individuals served on its Board of Directors:

   James E. Rush, Chairperson
     PALINET
   Michael J. Mellinger, Vice Chair/Chair-elect
     Data Research Associates
   Paul Evan Peters, Immediate Past Chairperson
     New York Public Library
   Heike Kordish, Treasurer
     New York Public Library
   Patricia R. Harris, Executive Director
     National Information Standards Organization

   Directors-Representing Libraries
   Lois Ann Colaianni
     National Library of Medicine
   Susan Vita
     Library of Congress
   Shirley Kistler Baker
     Washington University

   Directors-Representing Information Services
   Lois Granick
     American Psychological Association
   Michael J. McGill
     Ameritech
   Wilhelm Bartenbach
     Engineering Information, Inc.

   Directors-Representing Publishing
   Peter J. Paulson
     OCLC/Forest Press
   Constance U. Greaser
     The RAND Corporation
   Marjorie Hlava
     Access Innovations, Inc.