home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Danny Amor's Online Library
/
Danny Amor's Online Library - Volume 1.iso
/
bbs
/
society
/
society.lha
/
PUB
/
isoc_news
/
1-3
/
n-1-3-020.20.2a
< prev
next >
Wrap
Text File
|
1995-07-21
|
4KB
|
66 lines
N-1-3-020.20.2, "A Progress Report on the Implementation of the Z39.50
Information Retrieval Protocol", by Clifford A. Lynch*,
<calur@uccmvsa.ucop.edu>
The Z39.50 information retrieval protocol became a US national
standard in 1988; it was developed by the National Information
Standards Organization (NISO), the ANSI-accredited standards writing
body serving the library, information services and publishing
communities. The 1988 version saw little implementation, though it
was used (in modified form) as the basis for the current version of
the Wide Area Information Server (WAIS) system. Since 1988, work has
been underway which culminated in the completion of version 2 of the
standard, which was adopted in early 1992 and will be published
shortly. Version 2 of the standard makes use of ASN.1 as an encoding
mechanism, adds considerable functionality, and fixes a number of
problems with the original standard. There is also an international
analog (ISO 10162 and 10163) which is a subset of Version 2 of the US
standard; this has reached IS status.
A number of interoperable implementations of Version 2 of Z39.50 are
now operational on the internet, with more on the way. While the
Z39.50 standard was done within the OSI framework, the vast majority
of the current implementations run over TCP/IP; the approach is to
layer the Z39.50 application layer directly on top of a TCP/IP
connection using TCP port 210 to contact the server. Currently
running servers include bibliographic databases at the University of
California (the MELVYL system at the office of the president and a
campus-wide information server at UC Berkeley), Pennsylvania State
University (the LIAS catalog), OCLC, the Research Libraries
Information Network, and AT&T Bell Labs library. A number of these
organizations also have running clients, as well as Dartmouth
University and several library automation vendors, including Data
Research Associates, NOTIS, Gaylord, and Innovative Interfaces. A
number of other implementations are expected to be on the net within
the next few months; it is hoped that these will include a new version
of WAIS upgraded to support Z39.50 Version 2. A working group called
the Z39.50 Interoperability Testbed operating under the auspices of
the Coalition for Networked Information is seeking both to further
interoperability among the developing implementations and to extend
implementations to support additional nonbibliographic classes of
information, such as fulltext documents and images.
There is an active implementor's group which maintains a LISTSERV
called Z3950IW at NERVM.BITNET; this group is meeting quarterly to
move forward on the development of Version 3 of the standard, which
will add a great deal of additional functionality. It is hoped that
Version 3 will be ready for ballot sometime in 1993. One of the
extensions targeted for version 3 is a facility called EXPLAIN, which
permits a server to describe itself and its contents to clients and
eliminates much of the current need for outside of protocol agreements
between clients and servers in order to achieve interoperation.
Z39.50 is viewed as a basic building block for the development of
networked information servers; it provides a common search syntax and
protocol control structure allowing clients to search information
resources on the network and retrieve results from these servers
without regard to the specific database structures in which
information is stored.
*Director, Library Automation University of California Office of the
President