D-Lib Magazine
The Magazine of the Digital Library Forum
August 1995

Clips and Pointers

Version 3 of the NISO Z39.50 Computer-Computer Information Retrieval Protocol. Balloting has been completed on version 3 of the Z39.50 computer-to-computer information retrieval protocol developed under the auspices of the National Information Standards Organization (NISO). NISO will publish the standard formally later this year, but the final text is available through the Z39.50 maintenance agency's web site at the Library of Congress (http://lcweb.loc.gov/z3950/agency) along with the text of version 2 of the standard and other information valuable to implementors. Those interested in Z39.50 might also wish to explore the site at http://www.ds.internic.net/z3950/z3950.html, which contains a large number of pointers to documents related to Z39.50 as well as several software offerings. (Contributed by Clifford A. Lynch, University of California)

The Netlib repository (http://www.netlib.org/) is a collection of high-quality mathematical software, most of it freely available. In addition, the repository contains other material of interest to the scientific computing community, including software documentation, performance data, and technical papers and reports. Netlib began operation in 1985 and currently receives several thousand requests per day. (Contributed by Shirley Browne, University of Tennessee)

The National HPCC Software Exchange (NHSE) (http://www.netlib.org/nhse/) is an Internet-accessible resource that facilitates the exchange of software and information among research and computational scientists involved with the High Performance Computing and Communications (HPCC) program. The NHSE promotes the development of discipline-oriented software and document repositories, and encourages the use of such repositories by the high performance computing research community and by industrial and educational users. The NHSE home page provides a single entry point and search interface to the distributed collection of participating sites. (Contributed by Shirley Browne, University of Tennessee)

The Center for Intelligent Information Retrieval (CIIR), University of Massachusetts at Amherst, conducts basic research and technology transfer projects in text-based information systems with support from public agencies and private corporations. In addition to descriptions of their research projects, see their demonstration projects using INQUERY, based on material from the Library of Congress, Department of Commerce, and Lotus Customer Support (http://ciir.cs.umass.edu/index.html).

Uniform Resource Identifiers: An Explosion of Activity. The Uniform Resource Identifier Working Group of the Internet Engineering Task Force exploded! Not quite literally, but in the course of trying to organize the multiple efforts that were going on in the working group, it was decided to create several new groups that could focus more tightly on separate tasks. There are five separate areas of action: (1) reviewing the current URL documents before advancing them along the IETF standards track; (2) developing a standard for the syntax of Uniform Resource Names (URNs) and distribution of name assignment authority; (3) development and evaluation of a variety of protocols for looking up information and resources associated with a URN; (4) defining a standard for the syntax and structure for description of Internet resources (Uniform Resource Characteristics, or URCs); and (5) defining a well-known tag set (schema) for Internet resources. Mailing lists have been established for some of these efforts already. Although the URI working group has formally been closed down, the mailing list (uri@bunyip.com) will remain open for discussion of general URI issues. Proposals that have sufficient constituency and focus will be submitted for approval as new working groups of the IETF. (Contributed by Larry Masinter, Xerox Palo Alto Research Center)

ERCIM Fellowship Call. The European Research Consortium for Informatics and Mathematics (ERCIM) has issued guidelines for its 18-month fellowships for young scientists, who are engaged in digital library research. Areas of interest include: information and knowledge systems; interactive and multimedia systems; computer graphics, image processing, vision and visualization; distributed systems; authorization, authentication, security, and data protection; and copyright. In addition to approval of the research topic, applicants must meet the following criteria: have a Ph.D. degree, or be in the last year of the thesis work with an outstanding academic record and with a thesis topic related to digital libraries research; be fluent in English, not be (former) employees of the member institutions of ERCIM, and be discharged or have a deferment from military service. The full text of the announcement details areas of research and is available on ERCIM's home page (http://www-ercim.inria.fr/ ).

NCSTRL: Creating a unified digital library for computer science technical reports. The past few years have seen a number of experimental digital libraries for computer science technical reports, including UCSTRI, WATERS, and Dienst. A new project, the National Computer Science Technical Report Library (NCSTRL), combines the efforts of the latter two to create a single unified library. NCSTRL will benefit readers, who will see a large collection that hides any underlying differences; authors, who will reach a wider audience than before; and departments, which will save costs of printing and mailing technical reports (TRs).

NCSTRL builds upon technologies developed in the ARPA-sponsored Computer Science Technical Reports (CS-TR) (http://www.cnri.reston.va.us/home/cstr.html) project, Dienst (http://cs-tr.cs.cornell.edu/), and the NSF-sponsored WATERS project (http://www.cs.odu.edu/WATERS/WATERS-GS.html). The resulting architecture is both powerful and easy to install -- characteristics essential to achieving widespread participation in a distributed library of computer science technical reports.

The NCSTRL software will come in two versions, Standard and Lite. NCSTRL Standard sites will run a Dienst server, providing indexing, storage, and user interface services for documents at that site. A Dienst server provides a number of high level services for the site including multiple document formats, rapid browsing, and image zooming. Sites with fewer resources will run NCSTRL Lite, which stores documents in a single format on an FTP server while relying upon a dedicated central site to provide indexing and user interface services for the Lite sites. The central site will run a Dienst server, thus providing seamless access to the entire collection, regardless of the kind of server a site runs.

NCSTRL development is being coordinated by the D-LIB-based NCSTRL Working Group, which includes members from CNRI, Cornell, Stanford, the University of Virginia, Virginia Tech, and Xerox. The initial phase of NCSTRL is planned to begin operation on November 1, 1995. Later phases will migrate the technology from the Dienst architecture to the emerging digital library infrastructure. For more information, visit our Web site, http://willow.tc.cornell.edu/ncstrl.html. (Contributed by James Davis and Carl Lagoze) {Revised by Editor, August 17, 1995}

The Naval Research Laboratory (NRL) Library and the American Physical Society (APS) are collaborating on TORPEDO (The Optical Retrieval Project: Electronic Documents Online), a multi-staged project that embeds an experimental project in electronic storage and dissemination of APS publications (Physical Review Letters and Physical Review E) and NRL technical reports in a longer term study of how scientists use electronically accessible materials in their research. Roderick Atkinson and Laurie Stackpole describe the project more fully in their contribution to The Public-Access Computer Systems Review (http://info.lib.uh.edu/pr/v6/n3/atki6n3.html). The project maintains a home page at http://infonext.nrl.navy.mil/htbin/torpedo.pl.

The President's Information Infrastructure Task Force and the Council on Competitiveness are co-sponsoring the National Information Infrastructure Virtual Library (http://nii.nist.gov/nii.html). This web site, which is supported by the Office of Enterprise Integration of the National Institute of Standards and Technology, directs users to current research and lists of related projects as well as to basic information on the National Information Infrastructure (NII) and the Global Information Infrastructure (GII).

The International Federation of Library Associations and Institutions (IFLA) maintains a home page at http://www.nlc-bnc.ca/ifla/, which leads to a new page devoted to digital library resources and projects (http://www.nlc-bnc.ca/ifla/services/diglib.htm). The program for IFLA's August meeting in Istanbul and a summary of projects for the Information Technology section are available in the June issue of the IT Review (http://www.nlc-bnc.ca/ifla/pubs/sections/sit/sititr25.htm).

The proceedings of Digital Libraries '95, which was held in Austin, June 11-13, 1995, are available at http://csdl.tamu.edu/DL95/. The material includes copies of papers, lists of attendees and sponsors, and indexes by author and keyword.

The Library of Congress and the University of Virginia Library have issued the Proceedings of the Seminar on Cataloging Digital Documents, October 12-14, 1994 (http://lcweb.loc.gov/catdir/semdigdocs/seminar.html). The site contains a summary of the seminar, photographs taken by University of Virginia staff, text of presentations, panel discussions, list of participants and agenda, notes from Library of Congress' staff, and comments on the action plan and on the on-line presentation of the seminar.

SilverPlatter Information has invited participation in theWorldwide Library Sampler Project (http://www.silverplatter.com/sampler). The project involves free Internet access to 1,000-record subsets 63 databases, and enables participants to offer feedback in a number of areas, including the overall system and individual databases.

IBM issues reports and other publications from the corporation's research community as the IBM Research CyberJournal (http://www.watson.ibm.com:8080/). The journal is searchable and supports known-item (by ID number) as well as subject search (by keyword). Items published within the last 30 days are listed in a separate file in reverse, chronological order.

D-Lib Forum |  D-Lib Magazine Contents Page