
Title: National Agric. Library Electronic Information Initiative
Document-date: Nov 1993
Author: C. Ditzler, C. Early, C. Weston
Organization: National Agricultural Library, USDA
Source: URL:ftp://cliff.nalusda.gov//pub/elec.init/eii-rpt.txt
Contact-name:Brian Norris
Contact-phone: 301-504-6778
Posting-date: 29 Mar 1994
Length: 6813 lines 336939 char



                 THE  ELECTRONIC  INFORMATION  INITIATIVE


                           PHASE 1 FINAL REPORT



                        A Key Success Factor in the
                            NAL Strategic Plan






                               November 1993





                               Submitted by:


             Carol Ditzler, Caroline Early and Claudia Weston




                      U.S. Department of Agriculture
                       National Agricultural Library
                        Beltsville, Maryland  20705







                             Table of Contents



Executive Summary - Statement of Committment - Summary Recommendations &
Introduction

Chapter 1 - Evaluation Study

Chapter 2 - AGRICOLA On-line at NAL

Chapter 3 - Electronic Journals

Chapter 4 - CD-ROM

Chapter 5 - Electronic Research Data Bases

Chapter 6 - Electronic Versions of NAL Publications

Chapter 7 - Preservation Issues for Optical Media

Chapter 8 - Recommendations for Specific Actions

Bibliography

Appendices     -    Not Available Electronically








                             Executive Summary

To continue the National Agricultural Library's national leadership role in
the production, management and dissemination of agricultural information,
Director Joseph Howard launched The Electronic Information Initiative in
December 1992.  The purpose of the Initiative is to research, plan and
implement a systematic program of managing data in electronic form.  Phase I
of the Initiative consisted of an examination of the issues associated with
the Library's ability to manage electronically created and stored information.


A Steering Committee, consisting of two Branch Heads and the Library System
Coordinator, was charged with enlisting a broad-based group of NAL staff to
work on this project.  Three committees were established and charged to
proceed with a practical effort to "provide recommendations for changing
policies and procedures related to the acquisition, processing, access,
dissemination, collection maintenance and preservation" of:

--   Electronic journals

--   Original texts and reproductions of print documents disseminated on CD-
     ROM

--   Electronic research data bases.

Two additional committees were charged with examining and documenting:

--   The impact of converting NAL print publications to electronic format

--   The preservation issues associated with optical/electronic media (CD-
     ROM).

The committees gathered policy and procedural statements, identified
background material, reviewed the current literature,  brainstormed to
identify questions which needed answers and scenarios they might face in the
future,  proposed answers to as many questions as possible and finally, made
both broad and specific recommendations for actions to be taken by the
Library.  The most important recommendation is that the Electronic Information
Initiative Statement of Commitment be approved.  Committing to become an
"electronic library"  will strongly position NAL to meet the challenges of the
transition from providing information services using a traditional, print
collection to providing full electronic access to information.

Included in the Report are sections which address two additional  key
components of the Electronic Information Initiative, namely, the Evaluation
Study Developing Structures for Managing Electronic Information in
Agriculture,  and a proposal to make AGRICOLA available online at NAL and
accessible via the Internet.  This Report does not address the specific costs
associated with the recommended actions, rather, the recommendations are
arranged according to a suggested timeframe.   Flexibility has been designed
in to the recommendations to allow implementation of as many specific actions;
as possible.











                 THE  ELECTRONIC  INFORMATION  INITIATIVE

                          STATEMENT of COMMITMENT


The world of information management is changing daily.  The current paper-
based information delivery system is inadequate to keep pace with the needs of
the modern agriculturalist.  Increasingly,  information is produced in
digitized form, and with recent telecommunications innovations and the
Internet, the resources available to the computer literate researcher are
expanding exponentially.  The length of time expended for traditional
publication, procurement, physical handling, indexing and document delivery
processes is no longer acceptable.  Transition from the traditional print
collection to the "electronic library" will necessitate a shift from a
production mode to a facilitator mode of service. 

The National Agricultural Library (NAL) has a responsibility to provide
leadership to the increasingly networked agricultural information community.  
In its role as a national library,  NAL has an obligation to manage
information, promote standardized formats, repackage essential information for
mass distribution and mediate electronic access to information irrespective of
where the data reside.  In addition, NAL is an intermediary with a large
client-base that can facilitate USDA research initiatives and enhance the
dissemination of these research results to the world.

As knowledge takes on new forms, library staff must learn new skills and
develop new storage/dissemination models and systems suitable to meet the
requirements of the electronic information age.  The NAL is creating a
planning document for a systematic program of managing data in electronic form
and establishing strategies for collecting, storing and distributing U.S.
agricultural information in electronic form.



To demonstrate its commitment to meeting the challenges of becoming an
"electronic library", on January 1, 1995,  the National Agricultural Library
will designate electronic information the preferred medium.










                          Summary Recommendations


I.   Adopt The Electronic Information Initiative Statement of Commitment.

     A.   Establish it as one of the driving forces in budgeting, decision
          making, allocation of staff, setting priorities, etc., throughout
          the Library.

     B.   Widely publicize it throughout NAL, USDA and the Land-Grant and
          world-wide agricultural communities.
     

II.  Provide NAL Staff and Patrons with State-of-the-Art Electronic Access to
     Electronic Information   

     A.   Make information available via the Internet, NALNET, a CD-ROM
          Network, etc., using WAIS, Gopher, Veronica, etc.

     B.   Implement networking as the keystone to seamless access to
          information.


III. Add Staff and Funds to Acquire, Process and Make Available Electronic
     Data.


IV.  Make AGRICOLA Available Online at NAL.


V.   Reallocate for FY 94, a portion of the NAL budget and each Branch budget
     which must be used to move toward implementing the Electronic
     Information Initiative Statement of Commitment, until funds become
     available in FY 95.


VI.  Establish a Permanent Steering Committee on Electronic Information
     Issues, consisting of three Branch Heads, which would:

     A.   Create a plan for implementing the recommendations in this Report.

     B.   Keep the Initiative on track and moving forward and serve as
          consultants to all NAL managers regarding electronic issues.
          
     C.   Keep abreast of the changing technologies and their potential
          applications.


VII. Establish Cooperative Efforts with Other Interested Agencies and
     Libraries to:
     
     A.   Meet its obligation as a National Library to manage information,
          promote standardized formats, repackage essential information for
          mass distribution and electronically mediate access to information
          irrespective of where the data reside.

     B.   Advise USDA regarding publication standards (i.e., SGML).

          1.   Avoid duplication of effort.

          2.   Minimize costs.

          3.   Increase institutional influence on a variety of issues
               including licensing, copyright and other joint concerns.

     C.   Implement a preservation/archiving plan. 


VIII.     Establish a Variety of Permanent and Temporary Committees to Continue
          the Work of the Five Committees:

     A.   Electronic Media Collection Committee (permanent, broad-based, to
          approve the selection and acquisition of electronic media,
          recommend weeding or retention of comparable print titles,
          establish storage, leasing and processing policies and procedures,
          and assign processing priority and preservation status to each
          electronic format title).
     
     B.   Committee to investigate use of SGML to produce and acquire
          publications (ad-hoc, to determine the software NAL should use to
          produce its publications using SGML).
     
     C.   Preservation Committee (permanent, broad-based, to continue to
          study and evaluate preservation issues, keep abreast of both new
          technologies and the National Preservation Program for
          Agricultural Literature, and develop a program for monitoring and
          refreshing digital data, etc.).
     
     D.   Standards Committee (permanent, broad-based, to establish
          standards for equipment and internal electronic media production).

     

IX.  Review and Revise Policy Statements and Operating Procedures (both
     internal and external) to:
     
     A.   Reflect NAL's preference for the electronic version over print or
          other versions.
     
     B.   Require USDA to deposit copies of electronic documents, data bases
          and documentation at the NAL and to report publication of
          electronic media including electronic journals and newsletters
          (similar to the current requirement to deposit copies of print
          products).

     C.   Review and revise all NAL operating procedures, document changes
          and distribute widely.
     
     
X.   Establish, Equip and Staff an Electronic Media Center for Access to
     Electronic Media.


XI.  Establish a Library-wide Electronic Information Training Program to
     Include:

     A.   Formal and informal training for all NAL staff.

     B.   Continuing efforts to make staff and patrons aware of the latest
          developments in information provision systems and all aspects of
          electronic media, the electronic library, etc.

     C.   New Technology Seminars, etc., open to all NAL staff.
     

XII. Provide for the Expertise in the Library of Someone who can Advise NAL
     Regarding Licensing, Copyright, and Patents.


XIII.     Convert the NAL Print Publications Program to Electronic Media.


XIX. As Part of the Requirement to Provide for the Archival Storage of USDA
     and NAL Publications, Establish a Separate, Climate Controlled, Secure
     Location to House and Service the Archival Copies of all Electronic
     Media.

XX.  Develop and Enhance Policies, Procedures and Practices which Facilitate
     the Preservation of Electronic Media as part of the NAL Preservation
     Plan.

     A.   Design and implement a disaster plan for NAL.

     B.   Develop and implement practices that enhance the security of the
          electronic collection.

     C.   Immediately implement short range storage and use practices such
          as caddies that will prolong the lifespan of electronic media and
          protect them.

     D.   Develop, support and use production standards for electronic media
          that will permit preservation and access well into the future.


XXI. Incorporate the Work of these Committees into NAL's Strategic Planning
     Efforts.










                               Introduction

NAL's current paper-based information document delivery system which furnishes
articles cited and indexed in AGRICOLA and other data bases, is becoming
increasingly inadequate to keep pace with the requirements and computer
capabilities of the typical laboratory scientist or policy maker.  The spate
of information being produced in digital form, coupled with recent
telecommunications innovations and the dominance of the Internet, have
combined to vastly increase the resources available to agricultural
researchers and scientists.  Compared to the virtually instant electronic
transmission of information via the Internet, the considerable time required
to provide information through the traditional methods -- involving
publishing, procuring, physical handling and document delivery of paper
products -- is rapidly rendering the traditional model obsolete.  Thus, NAL's
need to embrace this already emerging reality termed "library without walls"
or "electronic library" dictates an immediate shift from providing information
services based on printed materials to methods in which NAL will facilitate
users in obtaining information electronically.

Core to all of the NAL programs has been the creation and advancement of the
AGRICOLA data base, which since 1970 has been the principal source of
agricultural information for practitioners and researchers worldwide.  With
over three million records, AGRICOLA citations and indexing records correspond
to the most important agricultural publications from around the world.  It is
not possible to overstate the importance of the information contained in the
AGRICOLA data base to the scientists and researchers of ARS, USDA, the land-
grant universities and other institutions nationwide.

As a national library, NAL is charged with the duty to manage information,
promote standardized formats, repackage essential information for mass
distribution, and electronically mediate access to information irrespective of
where the data reside.  In addition, NAL is an institution with a large user
base capable of facilitating USDA research initiatives and enhancing the
dissemination of the research results to the world.

With the implementation of the Electronic Information Initiative (EII), NAL
will be established as the central node in an electronic system, providing
agricultural information by directly linking users with the information
producers.  Further, this approach complies fully with the spirit and
intention of the revised Circular A-130.  The thrust of the initiative is to
implement a program aimed at achieving an effective transition from providing
information services using a traditional, print collection to achieving the
full capacity to provide electronic access to information via the Internet as
part of the National Research and Education Network (NREN).

The EII will create the infrastructure necessary to ensure that NAL will have
the capacity to function in this newly created environment, in which the
instantaneous transmission of information worldwide is fast evolving as the
standard.  The necessary infrastructure will include the design and
installation of an internal network fully compatible with the Internet. 
Library users will be provided with multiple use workstations with the
capability to access and process data in a wide variety of electronic formats. 
Increased computer storage will be acquired to make USDA-produced data as well
as all agricultural data available on the Internet.  A computer system within
NAL will be developed to provide access to critical data bases (e.g..,
AGRICOLA, genome data bases, et al.) for use by ARS and USDA patrons, other
Federal agencies, the land-grant institutions, farmers, ranchers, consumers
and other members of the general public.  Primary connectivity will be
achieved via the Internet as part of NREN.  Extensive, intensive training in
the use of new technologies related to this project will be necessary.  







                      CHAPTER 1  -  Evaluation Study 

               Developing Structures for Managing Electronic

                         Information in Agriculture

                     Proposal - Partially Funded FY 93

Problem

There is a growing demand on government to make its information resources
available to researchers, educators, and the public at large.  USDA offers no
central repository through which materials in electronic form can be stored
and made easily available for dissemination to the public.  In addition, more
and more scientific and technical print materials are being made available in
electronic format.  These electronic materials are a small but rapidly growing
component of NAL's holdings.  NAL needs to create an appropriate
infrastructure to accommodate the storage and dissemination of these
materials.  Much evaluative work is needed to make this transformation
quickly, effectively, and efficiently.  The problem this evaluation study will
address is how to optimize the availability of this electronic information to
the agricultural community.

Objective

Over the past two years, the Internet has seen an explosion in the number and
types of information resources and services.  Over the same period, the number
of people with access to the network has grown correspondingly.  This study
will evaluate alternatives for the collection, control, storage, and
distribution of U.S. agricultural information in electronic form through
Internet and other vehicles.  Proposed solutions will accommodate both local
and remote users with varying skills and equipment.

Approach

A number of electronic information provision models will be evaluated and a
taxonomy of electronic materials and services will be developed.  USDA
materials will be the primary focus of this study.  One of the proposed
evaluative documents will be the Extension Review, a serial publication
produced by the U.S. Department of Agriculture Extension Service.  This study
will also examine the use of such publishing standards as SGML (Standard
Generalized Mark-up Language) in the production of electronic information and
initiate a planning process whereby such standards can be integrated into USDA
publication activities.  A working prototype information collection, storage,
and dissemination system will be produced and evaluated by a user test base. 
Some computer equipment will be required to develop and evaluate this working
prototype.

Establishing effective systems for managing electronic information is a cross-
cutting issue affecting all agencies within USDA.  As part of the evaluation
study, NAL will establish an Interagency Oversight Committee to provide
guidance and advice throughout the project cycle to ensure that proposed
solutions can be integrated into Departmental systems structures. 
Representation on the Committee will include a broad mix of agencies (i.e. S&E
agencies; ERS; FSIS; OPA; OIRM) whose data and interest in electronic
information issues are characteristic of USDA-wide agency interests.  


Results

This study will produce alternatives for the large-scale management and
distribution of electronic information.   These alternatives will be used to
develop an infrastructure for the future dissemination of agricultural
information and are expected to further USDA initiatives to streamline its
automated systems.


Proposed Budget

Evaluate existing approaches       $13,000
Develop taxonomy of materials      $ 8,000
Acquire prototype media            $25,000   
Create prototype information system     $87,000
Conduct survey/evaluation          $ 8,000
Evaluative Computer Equipment      $25,000

           Total Requested         $166,000



Project Schedule

Evaluate existing approaches  January - March 1993
Develop taxonomy of materials January - March 1993
Acquire prototype media       April 1993     
Create prototype info system  April 1993 - July 1994
Conduct survey/evaluation     July 1994 - October 1994
Submit evaluation report December 1994



Organizations to Receive the Final Report

1.   Council on Library Resources
2.   Library of Congress
3.   National Library of Medicine
4.   Association of Research Libraries
5.   United States Agricultural Information Network



Proposed Cooperators

Iowa State University         USDA Office of Public Affairs

USDA Extension Service        Interagency Oversight Committee



Principal Agency Contact (ADODR)

PAMELA Q.J. ANDRE
National Agricultural Library

                                                       November 1992



                               Status Report

Background:

In 1992, the NAL began exploring the issues surrounding systematic collection
and dissemination of electronic information.  As a result of these
explorations and with the encouragement of USDA administrators, NAL Director
Joseph Howard launched the Electronic Information Initiative.  The purpose of
this Initiative was to plan and implement a systematic program of managing
data in electronic form.  In partial fulfillment of this goal, an internal NAL
Steering Committee was formed and charged with enlisting NAL staff to address
the issues associated with information stored on electronic media.  Committees
were established and instructed to proceed with a practical effort to "provide
recommendations for changing policies and procedures related to the
acqusition, processing, access, dissemination, collection maintenance and
preservation" of: electronic journals, full text documents on CD-ROM, and
electronic research databases.  The primary focus would be USDA-produced
materials.



Status

To assist in the evaluation of existing information models and the
implementation of a systematic program for managing and disseminating
electronic information, an evaluation study entitled 'Developing structures
for managing electronic information in agriculture' was proposed and funded. 
Based on input from the various Electronic Information Committees, select
items were identified as candidates for testing recommendations and refining
processing and dissemination techniques.  These items include the Journal of
Extension, Arabidopsis thaliana Data Base (AAtDB), and CGIAR CD-ROM. The
Extension Review had initially been selected as a candidate for evaluation,
however, due to publishing uncertainties it was not included in the initial
test group.  A brief description of each of the test titles follows:

Journal of Extension

The Journal of Extension is a journal produced by the Cooperative Extension
System, University of Wisconsin.  It is available in both print and electronic
form.  Upon subscription to the electronic version, a table of contents
message is sent automatically to the subscriber; the subscriber is then able
to request via e-mail the full text of one or more articles.

Arabidopsis thaliana Data Base (AAtDB)

The AAtDB is a genome database for the model organism Arabidopsis thaliana. 
The Massachusetts General Hospital was funded by NAL to collect and pull
together information relevant to the Arabidopsis database.  Staff at the Plant
Genome Data and Information Center receive advance notice from Mass General
for the date of new releases of the database.  After the release is
advertised, the new data and software can be downloaded using standard FTP
procedures.

CGIAR CD-ROM

The CGIAR CD-ROM (Consultative Group on International Agricultural Research)
is a full-text CD-ROM containing documents from around the world.  Some of the
original print documents are held by NAL, others are not.  NAL has agreed to
undertake archival responsibility for this information.

An internal working group representative of the units within NAL has been
established and is charged with processing the electronic journal Journal of
Extension.  Once work has been completed on this title, another working group
will be formed to examine the AAtDB and then another to process the CGIAR CD-
ROM. Each of these working groups will build on the information and procedures
of the group it follows thereby strengthening the strategies for processing
and providing access to various types of electronic media.



 
Timeframe

Dec. 1992 - June 1993    Identify Issues and Prepare Broad Recommendations
July 1993 - Feb. 1994    Refine Recommendations Based on Test Media
March 1994 - Aug. 1994   Expand Testing and Begin Routine Implementation
Sept. 1994 - Dec. 1994   Prepare Final Report



Expenditures

Workstations for processing and disseminating electronic information:

     Quantity - 4
     Cost - 23,141.98



Workstations

Based on the input from various EII Committees, a portion of the 'Developing
Structures for Managing Electronic Information in Agriculture' evaluation
study funding will be spent on establishing workstations throughout the
Library.  These workstations are intended to be used for processing and/or
providing access to electronic materials.  Some of these workstations will be
connected directly to the HP for access to ISIS and, ultimately, all of these
workstations will be connected directly to the NAL backbone.  What follows is
a brief description of the equipment that has been funded, its location, and
its perceived uses.

Cataloging monitor

A dual page color monitor (or grayscale monitor) 

This monitor will be added to existing equipment to assist in simultaneously
accessing electronic documents and OCLC for cataloging purposes.

Indexing workstation

486 computer
Dual page color monitor (grayscale is another option)
8 mb of RAM
220 mb hard drive
2 floppy drives (1.2 and 1.4 mb)
DOS 5.0
Windows 3.1
Mouse
Dot matrix printer
Ethernet card

This workstation will be used for simultaneous access to electronic documents
and ISIS for indexing purposes.

DDSB 3rd floor workstation

486 computer
15" VGA color monitor
8 mb of RAM
220 mb hard drive
2 floppy drives (1.2 and 1.4 mb)
DOS 5.0
Windows 3.1
Mouse
Laser Printer
Ethernet card
Multiple CD-ROM drive [8]

This workstation will be used to access and disseminate electronic
publications.


Public workstation in reading room

486 computer
15" VGA color monitor
8 mb of RAM
220 mb hard drive
2 floppy drives (1.2 and 1.4 mb)
DOS 5.0
Windows 3.1
Mouse
Dot matrix printer
Ethernet card
Multiple CD-ROM drives [8]


This workstation will be used to provide public access to electronic
documents.  It is envisioned that 4 of the 8 drives will be used to access
AGRICOLA.





                  Agricultural Network Information Center

                               Status Report


A key component of the evaluation study is to establish a prototype
Agricultural Network Information Center (AgNIC) to experiment with the concept
of a distributed discipline-oriented Agricultural Information Center as
opposed to the technically-oriented centers established in the past few years
by the National Science Foundation (NSF).  Agriculture is a multi-dimensional
discipline with public (experiment stations, extension services, other
government offices) and private (producers, agribusiness, suppliers and
customers) components.  Networking is the enabling technology to share
information of interest among them.

The AgNIC is an innovative prototype involving the first use of affiliated
organizations with a common disciplinary interest, to share agricultural
information via the Internet, which is expected to be expanded to a wider
environment with significant ramifications for society.  Based on the long-
term symbiotic relationship between elements of USDA and the land-grant
universities and the fact that the universities already have in place the
infrastructure for Internet connectivity as well as considerable experience in
accessing the Internet, the agricultural components of the land-grant
universities (the Agricultural Extension and Agricultural Experiment Stations)
were identified during the initial planning session as logical components with
which to start AgNIC.  A small group of organizations from the Science and
Education component that would have the most involvement with the land-grant
universities, and that would be within the scope of funding of the prototype, 
will also be targeted.

In March 1993, NAL chaired a planning session with representatives of Iowa
State University, the USDA Extension Service (ES), the National Science
Foundation and the commercial sector Internet network managers, AT&T Bell
Laboratories and General Atomics, to begin planning for the development and
implementation of the AgNIC prototype.  Iowa State University was tasked to
name an Information Specialist to work full-time on the AgNIC prototype.  The
initial tasks of drafting a White Paper which defines the concept, mission and
goals, organization structure and resources required is presently in progress.






                   CHAPTER 2  -  AGRICOLA On-line at NAL


Background

AGRICOLA is a machine-readable data base of bibliographic citations created by
the National Agricultural Library and its cooperators.  The scope of this data
base encompasses all aspects of agriculture and allied disciplines including
plant and animal sciences, forestry, entomology, soil and water resources,
agricultural economics, agricultural engineering, agricultural products, and
food and nutrition.  Auxiliary subjects that support NAL's Information Center
activities such as agricultural trade and marketing, rural information, and
biotechnology are also included.  

Researchers, scientists, educators, and information providers are among the
collection of users who rely upon the information found within the data base. 
Examples of the organizations with which the searchers of AGRICOLA are
affiliated include USDA, research and educational institutions (the land-grant
universities are heavy users of the data base), trade associations, and
private industry.  Some of these organizations are international in nature.  

To provide access to AGRICOLA, NAL sends a monthly magnetic tape to the U.S.
Department of Commerce's National Technical Information Service (NTIS) for
further distribution to its subscribers.  These subscribers then determine the
frequency with which their data base products are updated, what medium will be
offered, what search capabilities will be provided, and the associated costs
to the end user for accessing and retrieving information.

This method of distribution began in the early 1970s, when mitigating factors
forced NAL to decide to distribute magnetic tape copies of AGRICOLA to
commercial subscribers and to use the resulting commercial products rather
than to develop end user products within NAL.  At that time, computers were
very large, very expensive, and required highly-skilled computer specialists
to develop applications and to keep them running.  NAL was not in a  position
to even contemplate undertaking such an endeavor.

Times have changed: a)  Computer power has increased tremendously;  b)
Computer prices have dropped dramatically and c) Easy-to-use software is
available to greatly facilitate the development of data base applications.


Problems with the Current Method of Distribution

The only access the National Agricultural Library has to its own bibliographic
data base, AGRICOLA, is through commercial vendors.  There is one source for
print, one source for online access, and one source for CD-ROM.  If any of
these three commercial companies decide to cease its AGRICOLA products, NAL
and, for that matter, the world will have no equivalent alternative source for
agricultural information.  NAL also has no input into the prices being charged
end users for data base access nor any control over search capabilities or
retrieval options provided by these commercial companies.


Benefits to Having AGRICOLA Online at NAL

The greatest benefit to the agricultural community in making AGRICOLA
available through NAL would be the reduction in the cost to access and
retrieve information stored within the data base.  This would benefit NAL,
USDA libraries, USDA researchers and practitioners, land-grant universities,
1890 universities, and other public and private organizations both national
and international.

In addition to reducing costs, the production and dissemination of information
stored within AGRICOLA would be centralized, thereby providing NAL with
greater control over the display, retrievability, and modification of the data
stored within.


Users

NAL should mount AGRICOLA on a computer platform at NAL which can be connected
to Internet with no access restrictions; that would allow anyone, worldwide,
who has access to Internet to search the data base.  Initially NAL will
probably want to have more control over the user base, at least until the
completion of an evaluation of the computer system capabilities and the load
this application will place on them.


Security

Appropriate security measures will be taken to ensure the integrity of the
AGRICOLA file and also guarantee the integrity of other NAL systems. 


Impact on Current AGRICOLA Subscribers

The current domestic subscribers to the AGRICOLA tape service include:

     Oryx Press -- Printed Bibliography of Agriculture

     BRS, Dialog and OCLC -- Online commercial data base
          [BRS has notified us of their intention to cancel their
subscription.]


     SilverPlatter -- CD-ROM data base

     Clemson, Cornell, and Iowa State -- online campus networks


Putting AGRICOLA on Internet will probably have little effect on the printed
Bibliography of Agriculture.  

Because BRS, Dialog, and OCLC charge their users for access to AGRICOLA, any
users who had the capability to switch to NAL's free Internet access to
AGRICOLA would probably be tempted to do so, although there are a number of
reasons why not all would want to switch.  The online vendors would probably
lose revenue from decreased usage of their versions of AGRICOLA.

SilverPlatter would probably lose some subscriptions to the CD-ROM version
because many of their current customers also have Internet access.

The campus networks may or may not be affected because most universities do
not want their students to have unrestricted Internet access but the $10,000
annual charge (subscription and usage fees) and the computer resources
required may become too great a burden at some point.


Hardware/Software Options

A preliminary analysis was made of three options for putting the AGRICOLA file
up at NAL.

1.   Make AGRICOLA available through the Integrated System for Information
     Services (ISIS) Online Public Access Catalog (OPAC) software on the
     HP3000.

2.   Install the AGRICOLA file on a Sun workstation using Cuadra's STAR
     software.

3.   Use SilverPlatter's SPIRS software to access the contents of the
     SilverPlatter AGRICOLA CD-ROMS which are downloaded to run on magnetic
     drives for greater efficiency.


Three separate reports follow which contain discussions of benefits and
problems of each of the options.  A number of other options are also possible,
but these three represent the widest diversity in terms of probable
implementations at NAL.



Conclusions

The apparent best choice for further analysis is the option of putting
AGRICOLA on a Sun workstation (or similar UNIX machine) using STAR or an
equivalent software package for search and retrieval.

Having AGRICOLA on a separate system will allow far greater security than if
it were combined with other library functions on ISIS.  A UNIX system will
allow for far greater flexibility and connectivity than the DOS-based machine
that SPIRS would require.




                                 Option 1

                           ISIS (VTLS Software)

ISIS (Integrated System for Information Services) is an integrated library
system through which a number of library functions are performed.  Presently
these functions include:  acquisitions, serials control, the creation of item
records, authority control, and the provision of an online public access
catalog.  This online public access catalog reflects the machine readable
holdings of NAL, the USDA Agricultural Research Service Regional Research
Center Libraries, the National Arboretum and a subset of the indexing records
created by NAL's Indexing Branch and its cooperators (December 1989 -
present).  Within the near future, the functionality of this system will be
increased to include: online indexing, the production of the AGRICOLA sales
tape (ATPS), and circulation.  VTLS, Inc. is the vendor who supplied the
proprietary software through which these functions are made possible.


Capacity

NAL's implementation of this software runs on a Hewlett-Packard (HP)
3000/980/100 minicomputer.  The data which include 828,000 bibliographic
records, 1,300,000 authority records, and 124,000 item records currently
reside on 14 1.3 gigabyte and two 670 megabyte disc drives which are at 65%
capacity.  The estimated growth rate for bibliographic records is 80,000 per
year, item records - 23,000, and authority records - 90,000.  The CPU is
presently running at 20 percent capacity, but the impact of online indexing,
ATPS, and circulation is still unknown.

The ISIS data base consists of approximately 28 percent of the bibliographic
records found in AGRICOLA.  The largest gap is found in the coverage of
indexing records.  AGRICOLA contains indexing records created from 1970 to the
present, while ISIS contains those created from December 1989 to the present. 
This gap reflects a difference of 2,172,000 records.


Strengths

The primary purpose of ISIS is to function as an integrated library system. 
The strength of an integrated library system is in its ability to share data
between different functions.  This ability minimizes the need to re-key data
and permits users to obtain different types of information without having to
access a number of different systems or files.

ISIS is also a production system which has been in operation for over five
years.  This has led to increased expertise among NAL staff in its operation
and capabilities.  A new system would require additional training for systems
and support staff as well as in-house staff and external users who would use
the new system for its retrieval capabilities.

The licenses that are currently in effect for both the system software and the
VTLS library software cover an unlimited number of users.  In providing for an
unlimited number of users, no additional software licensing fees would be
required to increase access to ISIS.  The only 
additional cost would be for upgrading the hardware to support an increased
number of records.

 
Weaknesses

The strength of the various vendor supplied AGRICOLA data bases is in the
sophisticated search and retrieval features offered.  Although the VTLS
software supports standard searches such as author, title, call number, etc.,
the ability to develop sophisticated search algorithms is lacking.  This would
be a significant disadvantage to many of the present AGRICOLA users.

Related to the lack of sophisticated retrieval capabilities is the lack of
sophisticated report and statistics generators.  The VTLS software does offer
some report writing and statistical tracking capabilities, but this is an
acknowledged weakness within this system.  VTLS has begun to focus more
attention on these areas.   

An additional weakness which may impact on any 24-hour per day system is the
lack of trained computer staff to monitor system operations overnight.  The
present hours of ISIS availability are Monday through Friday 6:00 a.m. to 7:30
p.m. EST.  To increase the hours of system availability, additional funding
would be required to hire competent staff to trouble-shoot system
irregularities in the off-hours.



 
Risk

The risk of losing existing data is minimal due to the daily backups performed
and the continuous transaction logging.  Although a thorough risk analysis has
not been conducted, some obvious prospective concerns exist should the ISIS
system become inoperable.  ISIS is a production system as well as one that
offers end-user searching capability.  Examples of the difficulties which
would arise should the system become unavailable include:
     
1.   In-house staff would be forced to search vendor-supplied data bases, at
     a cost determined by the vendor, to retrieve NAL-produced bibliographic
     records.  

2.   Automated activities would revert back to manual processing  which in
     turn would produce a backlog of information which would need to be
     entered into an automated system once it became operational.  

3.   Information that was made available through the integrated system would
     once again reside in separate files in disparate locations.

Once online indexing and ATPS are fully implemented, the impact of system
downtime on NAL will become more severe.  At present, there is no contingency
plan for VTLS online indexing or ATPS should the system fail.  The present
production system is outdated and will no longer be supported.  Discussions
are currently underway to develop such a contingency plan.  


Security

The ISIS system currently takes advantage of HP's 3-level security and VTLS'
hardcoded and parameterized user level security.  The University of North
Carolina at Charlotte, one of the VTLS libraries that has its OPAC on the
Internet, relies solely on this password control to provide security to its
system.  This present form of security can be further strengthened by
implementing HP's file level access security via lockword. 

Other options for increasing the security of the system include encapsulating
the VTLS system and providing an intermediary system that would dictate the
user login.  Encapsulation prevents the user from accessing the operating
system.  Should a VTLS software error occur, the user is automatically logged
out of the system.  By using an intermediary system to perform the log-in, the
access level would be predetermined thereby limiting the capabilities of the
user.  Once a user is logged in as a certain type, this type cannot be changed
until the user logs out and logs back in again.  These two actions would
significantly increase the security of the ISIS system.


User Support

Two mechanisms currently exist for ISIS users to report technical
difficulties.  The first of these is the ISIS support line which consists of a
local phone number users can access to report problems.  When an ISIS staff
member is unavailable to answer the line, the user is able to leave a recorded
message on an answering machine.  This answering machine is checked
periodically by ISIS staff during the 8:00am - 4:30pm work day.  The second
method through which users can report difficulties is by sending an electronic
message to the ISIS support e-mail address (ISISHELP@NALUSDA.GOV).  This
mailbox is checked regularly and an e-mail sent acknowledging receipt of the
message.  Further support should be provided to address non-technical user
concerns and questions.


Documentation

A number of HP- and VTLS-generated manuals and documents exist which describe
the nuances of running the ISIS system.  Much of the end-user documentation,
however, is in the process of being re-written to accommodate the latest
release of the VTLS software (VTLS-92).  Initially, activities will focus on
refining and developing print documentation.  Once progress has been made in
this area, the on-line documentation will be examined.  The VTLS software
offers a great deal of flexibility in the design and content of the various
HELP features.



AGRICOLA VS. ISIS OPAC

If AGRICOLA is going to be brought up in-house, some clear distinctions need
to be made between AGRICOLA and the Integrated System for Information Services
(ISIS) Online Public Access Catalog (OPAC) file.

Because ISIS is an integrated system which provides automated library
functions for NAL, the only records which should be included in ISIS are those
which represent materials which NAL processes.  If an item is not in the NAL
collection or in the collection of an NAL "branch" library (e.g., DCRC,
Arboretum), its bibliographic record should not be in ISIS.  Records for items
held elsewhere (e.g., ARS Regional Libraries, remote-based electronic
journals) should be included in AGRICOLA but not in ISIS.

The search and retrieval software for the AGRICOLA file will be far superior
to the OPAC search capabilities, but the ISIS software will include software
for bibliographic record input, circulation, serials control, acquisitions,
etc. 








                                 Option 2 

              AGRICOLA on SUN Server Under Cuadra's STAR DBMS

It is proposed that AGRICOLA be mounted on an NAL in-house computer system,
using the Cuadra STAR data base management software as search engine and
report generator for online queries by NAL personnel and collaborators.  NAL
already owns limited licenses for the two software products necessary to
convert, mount, index, and query MARC-format records, and NAL personnel have
experience in their use.


Software Considerations

Solaris

Solaris Operating System (Sun Microsystems) is an implementation of UNIX
system V.  It is a robust operating system with excellent built in utilities. 
It has good security features such as login control with passwording,
individual file and directory permissions, auditing and authentication.  It
conforms to level C2 of the National Computer Security Center's standards set
out in "The Orange Book".


STAR

The Cuadra STAR software, two copies of which are presently licensed to NAL
(one Sun workstation and one Alpha Micro supermicrocomputer), is a full-
featured text management system, with sophisticated Boolean search
capabilities and excellent report generation features.  Major functions are
Database Definition, Data Entry, Search and Report, Global Change (batch
operations on a number of selected records), and System Management.

Searching is rapid, using indexes that can be built on as many or as few data
fields as desired.  Word, occurrence (entire string), and full proximity
indexing are options that may be selected by the data base administrator.  The
number of records in a data base is limited only by the amount of hardware
disk storage; any single record can contain up to about two million
characters.  Input can be validated by a number of rule types, as well as
against authority files (lookup tables).

Data base security may be implemented by a system of user IDs and passwords
that are additional to those used by the operating system.  Passwords can also
be used to limit access by data base and function.  Input screens as well as
output formats (screen displays and print formats) can be customized to suit
the preferences and sophistication of the users.



The STAR software runs on System V UNIX on many platforms.  It has been
certified on Sun Microsystems' SUNOS and on Solaris operating systems.  This
DBMS has been used successfully in a number of organizations with multiple
concurrent users.  It has been tested on a data base containing one million
records, and retrieval was acceptable rapid.  STAR has most often been used
for in-house data bases with local users, but some installations have used it
with success for public-access data bases.  STAR excels in data base
production, because data can be input in a variety of ways, and then output
can be easily adapted for uploading to other software.

Cuadra Associates is unable at this time to certify that STAR will meet the
needs of a large number of concurrent users on a data base the size of
AGRICOLA, but they would be interested in a pilot project with NAL to do
performance benchmarking and testing to see if the product meets our needs for
speed, capabilities, and ease of use.


MARC Conversion Software

NAL owns the software necessary to convert records from and to MARC
communications format.  This is another Cuadra product that operates with the
STAR software.  NAL staff have tested it on AGRICOLA sale tapes, and
successfully created conversion files that were then loaded into a fully
searchable STAR data base.  Loading from tape, MARC conversion, and loading
and indexing of one month's worth of the converted AGRICOLA records into a
STAR data base takes one to three days, depending on the load on the system
and the total number of records on the tape.


 Other Considerations

Cuadra STAR software licenses are priced according to the number of concurrent
users.  NAL's STAR license on the Sun workstation is a seven year license. 
The MARC conversion programs (that are a part of the LIBCAT library automation
system) are already owned by NAL (Plant Genome Project) and no further
expenditure would be necessary except a normal maintenance/upgrade fee.


Hardware Considerations

Disk Storage

Disk storage has been estimated by calculating storage requirements of the
present three million plus AGRICOLA records plus 100-200 percent indexing
overhead.  AGRICOLA records on Dialog occupy about 12 GB at present.  As a
test, a sample set of 10,970 AGRICOLA records from the November 1992 sale tape
were converted and loaded into a heavily indexed STAR data base; the storage
figures for that test set are:



          Data base records                  13,313 K
          Indexes                       13,278 K
          System records                    403 K

          Total                              26,994 K, or 2.46 K/record

If all 3,005,539 AGRICOLA records with indexes occupied 2.46 K (a high figure
because of the older records which were much smaller) the total disk storage
required for a STAR implementation would be 7.4 GB.


SPARCserver

For adequate CPU throughput and random access memory on a machine running STAR
and AGRICOLA and servicing multiple users, NAL will probably need a multi-
processing server with a very fast processor.  Sun Microsystems has announced
a machine called the SPARCserver 1000 that is expandable to up to 8 50-MHz
SuperSPARC+ processors, 2 GB RAM, 12 SBus slots, and 100 BG of disk storage. 
A machine can be purchased with base configuration, and over time modules to
increase capacity can be added.

The machine and peripherals described above would probably suffice for the
first few years of operation, for they represent Sun Microsystems' newest and
fastest server technology, and the peripherals are reliable and reasonably
state-of-the-art.  The computer itself was designed with expandability in
mind: disk storage, processors, and memory modules could be added as needed
when capacity is exhausted.  Peripherals such as a SCSI terminal server to
which modems may be attached are low-cost and can be added if dial-up access
is desired.




                                 Option 3

                           Using SPIRS Software

It is proposed that AGRICOLA be mounted on an NAL in-house computer system,
using the SPIRS retrieval software that is being used on the CD-ROMs being
sold by SilverPlatter, Inc.  NAL personnel already have experience in its use.

The first issue was to try to identify some of the concerns associated with
using SPIRS on a PC with an interface to the Internet.  Based on discussions
with a network expert, sophisticated network operating system (NOS) software
such as Novell Netware will be needed in order to make the PC workstation talk
to the Internet node at acceptable speeds.   It would be necessary to ensure
the SPIRS PC would have the necessary hardware to act like a NetWare server
(or buy a server).  With such an NOS purchase there would be a benefit from
obtaining one of the third-party products like OnLine's OPTI-NET.  Use of
magnetic storage instead of CD-ROM drives would help speed of access, but the
costs would increase.  This is based on providing fast, multi-user searching. 
In addition,  a PC based CD-ROM system does not have a lot of security built
in, especially when considering both dial-in and Internet access. 

It is recommended that based on the cost and complexity of a PC-based system
with somewhat limited SPIRS search system and limited security, this Option
should not be adopted. 









                     CHAPTER 3  -  Electronic Journals

Summary

The Committee interpreted its original charge as follows:

1.   Identify questions and issues concerning electronic journals (e-
     journals) within NAL.

2.   Identify any existing policies and procedures needing change within NAL.

3.   Come to a consensus, if possible, regarding recommendations based on the
     issues identified, including, wherever applicable, cost and
     implementation timetables.

4.   Place experimental electronic journal records on ISIS in order to
     develop procedures for their processing.

Concurrently with the questions and issues phase, an educational process was
begun which consisted of identifying and reading pertinent articles,
subscribing to electronic discussion groups on Internet, and inviting other
members of the NAL staff to attend our committee meetings and share needed
knowledge.  One committee member took the opportunity while attending the VTLS
Users' Meeting at VPI to visit an expert who is involved in the University
Libraries' Scholarly Communications Project.  She returned with Powell, James,
"Scholarly Communications Quarterly Technical Report" 9/91-11/92 and 12/92-
1/93 which helped place VPI's electronic journal processing in the broader
context of other electronic experimentation taking place at VPI.  These
reports and other documents which proved useful to the committee are listed in
the Bibliography section. 

The Committee realized that some of the issues were common to several areas,
that is, the Acquisitions & Serials (ACQ/SER), Cataloging, Document Delivery,
Indexing,  and Reference as well as the Information Systems Division Committee
members all developed similar ideas and needs concerning training, networking,
and equipment.  These issues form the basis of the Recommendations which
follow. 


Recommendations

1.   Encourage all USDA agencies to report publication of electronic media
     including electronic journals and newsletters to NAL.

2.   Form a small committee including representatives from key functional
     areas of NAL to perform selection of electronic journals.  This
     committee should have final authority to choose e-journals suggested not
     only by Acq/Ser but by all interested NAL staff and patrons.  They will
     also decide the level of processing as outlined below.

3.   Use the following electronic journals for initial experimentation and
     documentation of procedures:  

     A.   Extension Review

     B.   Journal of Extension

     C.   Social Sciences in Forestry

     D.   Laboratory Primate Newsletter

4.   When an electronic journal is published in more than one format, choose
     the format which is most advantageous to NAL whether print, online, CD-
     ROM or diskette, until January 1, 1995 when the electronic version
     should be preferred.

5.   Establish the following short and long term processing levels.  They can
     be phased in as staffing levels, funds and equipment capabilities
     permit.

     A.   Short term - enter a bibliographic record in ISIS or AGRICOLA
          which provides remote access information.

     B.   Long term - enter a bibliographic record in ISIS with full text
          archived on an NAL computer.  Remote access information should be
          included.  

     C.   E-journals produced by USDA should always be archived on an NAL
          computer,  with the exception of newsletters.

     D.   E-journals which are state publications should receive second
          priority for archiving on a case-by-case basis.

     E.   All electronic journals within NAL's scope should receive a
          bibliographic record in ISIS/AGRICOLA even though it proves to be
          impractical or impossible to archive every title for various
          reasons.  This will apprise NAL patrons of the journal's existence
          and will provide remote access information.  Any title selected
          for indexing in AGRICOLA, however, should be considered for
          archiving by NAL.

     F.   Full CONSER level cataloging records should be provided for
          electronic journals just as they are for print journals.

6.   Multiple-use workstations should be connected to the NAL Ethernet in
     order to have direct access to full text of e-journals on an NAL
     computer.


7.   Off-site locations such as the DC Reference Center (DCRC) and ARS field
     libraries should be networked in order to have direct access.           

8.   Space should be made available on either the  HP, Cliff  or a new
     machine so that full text can be transferred and then retrieved using
     file transfer protocol (ftp).  

9.   Prepare an ftp site on a UNIX based system.  A UNIX based system is
     recommended because it is designed for multi-user access of large files. 
     

10.  Prepare a user friendly interface for a new system.

11.  Additional staff support in Information Systems Division should be
     provided to prepare the interface.  Lack of such an interface will
     significantly affect staffing needs in the Public Services Division and
     will make cataloging and indexing procedures more time consuming and
     expensive.

12.  Create an Electronic Media Center and staff with at least one systems
     librarian.  This Center would be the service center for all NAL staff
     and patrons.

13.  Establish a training program for all NAL staff in the use of the
     technology necessary to the acquisition, processing and dissemination of
     electronic media.

14.  Implement dissemination of internal administrative memoranda via e-mail
     as a means of acquainting all NAL staff with what is becoming a basic
     method of communication.



Issues

Training

All staff who will deal in some capacity with electronic journals should
immediately receive introductory training in the use of Internet resources and
mechanisms such as file transfer protocol, telnet and gopher.  This type of
training is beginning to be more widely available.  As an example, a private
company has advertised a workshop called "Understanding the Internet" which
the company describes as being a half-day beginner's seminar on Internet
technologies.   The Association for Research Libraries also sponsors a series
of workshops concerning networking and the Internet at a fairly reasonable
cost.  The recent Computers in Libraries Conference was very helpful in
gaining basic knowledge of Internet resources.  Additionally,  the annual ALA
meeting includes several programs which concern electronic media. 



A long term practical training program for all staff should include at least:

1.   skills necessary for transferring full text to an NAL computer. 

2.   training in Internet resources as well as basic computer skills. 
     Document Delivery staff and contractors must have the expertise to
     access electronic journals archives whether they are at NAL or another
     repository.  They must then know how to either transfer the files to the
     requestor via electronic mail, download the file to floppy disk or print
     it out on paper.

3.   training in the use of new equipment, software and systems such as dual
     page monitors, Windows, and the coming Ethernet.


Many university libraries are now using electronic mail as a means of
disseminating internal administrative memoranda.  It is noteworthy that Gail
Macmillan of VPI in her article "Electronic Journals: Present and Future
Considerations" Serials Review 17 (4), 1991 said that problems with receipt of
e-journals were not anticipated since "all Libraries' staff has experience
receiving and forwarding electronic mail."



Networking and Archiving Capabilities

In the Cataloging and Indexing Branches it will be necessary to access the
full text of an e-journal in order to process it properly just as it is now
necessary to obtain full bibliographic information and full text of a title
which is in hard copy.  The only truly feasible manner by which to do this is
to have direct access to the full text.  With proper equipment such as Windows
and dual page monitors, Cataloging and Indexing staff can access the full text
in one system while simultaneously cataloging on OCLC or indexing on ISIS. 
Downloading to print so that cataloging and/or indexing can be performed is at
best impractical and in some cases virtually impossible.

The importance for all Public Service Areas (Reference, Information Centers
and Document Delivery) to be wired to the NAL network can not be overstated. 
Access to NAL's local area network by DCRC, ARS field libraries and other USDA
installations would, over a period of time, lighten the work load of
Reference, Information Centers and Document Delivery Services as researchers
learn to access electronic journals themselves using an interface developed
for a new machine.


Policies Requiring Change

The Electronic Journal Committee examined two documents with regard to current
policy as they pertain to this format.  They were: Departmental Regulation
1020-1 revised March 23, 1990 and Selection Policy No. 1, Machine Readable
Materials, 3/21/85. These two documents were studied with a view to
identifying areas requiring change.  Selection Policy No. 1 was examined in
conjunction with several similar selection policies from other libraries
provided by the Steering Committee.


I.   Departmental Regulation 1020-1, Subject: National Agricultural Library,
     March 23, 1990.

This Regulation does contain scattered references to electronic materials. 
For NAL to expect optimum cooperation from USDA agencies to supply either the
electronic media or the access information to NAL, the Regulation should be
worded more specifically. 

Paragraphs and subparagraphs which the committee believes should be reviewed
for this purpose are:

     5.   DEFINITIONS

     f.   Non-Expendable Published Material.

          Change to read: "Publications in the NAL collection, the
          collections of the agency libraries and information centers,
          reference aids, and all published materials not included under
          expendable published materials are classed as non-expendable,
          including electronic media."

     i.   Add ISIS - NAL's online public catalog

     j.   Add ALF  - NAL's electronic bulletin board


     9.   GENERAL POLICY

     b.   Publications.  

     1.   "Publications" can be construed by USDA agencies to mean only
          printed media.  The committee recommends that a sentence be added
          such as the following: "Publications" should be interpreted to
          include materials in electronic and audiovisual as well as printed
          formats.  One copy of all audiovisuals, software packages, CD-ROMs
          and any other electronic media  developed by USDA should be
          forwarded to NAL.  In addition,  access information to electronic
          journals and research data bases should also be provided to NAL.  

     2.   Change to read "Books or other published materials as well as
          audiovisuals and electronic media. . ."


     14.  ACQUISITION

          Nothing will need to be added here if the definition of
          publications has been expanded in 9. GENERAL POLICY.

     17.  FOREIGN EXCHANGES.

     a.   Exchange Agreements. 

          No change necessary if definition of publications has been
          expanded.


     19.  DISPOSITION OF LIBRARY MATERIALS

     No change necessary if definition of publications has been expanded.

     A time may come when nonexpendable materials offered to NAL by USDA
     agencies will include e-journal archives on tape, diskette or CD-ROM
     just as they now bring to NAL archives of bound printed journal titles. 
     (It is more likely, however, that only NAL or NAL in conjunction with
     the Office of Information Management (OIRM) will go to the trouble of
     archiving USDA e-journals.) 


II.  Selection Policy No. 1 - Machine Readable Materials 3/21/85

An initial recommendation: Rename the document to Electronic Media Selection
Policy.

As before, paragraphs and subparagraphs which should be reviewed for change
are asterisked and often have attendant recommendations.

Change the second sentence to read "In addition to materials to be physically
added to the collection, access information to online materials (e.g.,
electronic journals, newsletters, and other data bases both bibliographic and
numerical) are also covered under this policy."

"Materials or access to data bases needed for operational use by the library
units are not included in this statement."  This sentence can remain as it is,
the committee understanding it to mean software packages such as WordPerfect,
dBASE, etc. and data bases such as OCLC.


Selection Criteria:

A.   Change second sentence to read: "All types of electronic media such as
     magnetic tape, floppy disk, and CD-ROMs  may be collected as well as
     archives and/or access information to online data bases."

3.   There should be no time limit with regard to the archiving of USDA
     produced electronic journals.  In order to save computer space, perhaps
     they could be transferred to tape or CD-ROM after 3 - 5 years.  A time
     limit of current plus three previous years could be considered for non-
     USDA electronic journals.

5.   This paragraph should be omitted and USDA electronic materials such as
     e-journals should not be subject to this type of selection criteria. 
     Instead employ the criteria in Paragraph D.  Even in the case of USDA e-
     journals and other USDA electronic media, there may be instances where a
     bibliographic record giving access information will be all that is
     practical or desirable.


B. & C. 

The statements under these two categories are still valid, but USDA electronic
materials must be serviced regardless of such considerations as cost or
inconvenience.  

B. 1 Based on whether the Statement of Commitment is adopted, the format of
     choice may need to be changed. 
D.   A good paragraph in that it discusses electronic access vs physical
     custody.  

E.   This paragraph seems to contradict Paragraph A. - 5., but E should be
     the paragraph of choice, especially if it is worded in a manner similar
     to the following:

     "Electronic media whose contents are suitable for the collection,
     especially USDA materials, will be acquired even when problems of
     processing or access remain unsolved if no other form is available". 
     Again, however, there may be instances even with USDA material when
     physical custody is not practical or possible and access information
     only is provided in ISIS or AGRICOLA.   


Selection\Collection Development

Selection of e-journals should be the primary responsibility of the
Acquisitions & Serials Branch just as it is their responsibility to select
serials in print format.  Acq/Ser will maintain close liaison with Cataloging
and Indexing Branches regarding potential choices as well as with Public
Services Division and Information Systems Division.  This liaison should take
the form of a small committee.  This committee would decide not only which e-
journals will receive bibliographic records in ISIS but of those choices which
ones will be archived at NAL.  It is not recommended that the Collection
Development Committee be used for this purpose since the goal is to have a
very specific charge, thereby enabling the committee to make decisions in an
expeditious manner.   

USDA agencies should be required to report publication of electronic media,
including e-journals, to NAL, including bibliographic, access and primary
contact information, as well as hardware and software requirements.  It would
be preferable for all reports to come directly to NAL rather than indirectly
through OIRM.  First points of contact should be Acquisitions & Serials Branch
and Cataloging Branch.  Cataloging Branch has already developed forms on which
USDA agencies may describe their electronic records for input into a Catalog
of USDA Electronic Information.  The information fields within the forms are
MARC tagged to enable ease of entry into ISIS and they differ according to the
type of format being described such as CD-ROM, diskette or online data base.

The e-journal being considered must fall within NAL's subject scope with core
subjects receiving higher priority.

Those journals selected will receive processing, broadly speaking, in one of
two ways with all decisions, other than those involving USDA e-journals, being
made on a case-by-case basis:       

1.   Bibliographic records, which include access information, will be placed
     in ISIS and full text of the e-journal will be transferred to an NAL
     computer.

     a.   USDA publications such as Extension Review will be given first
          priority with regard to this type of processing, though USDA
          electronic newsletters will not be archived.  An example of the
          latter type of publication would be Extension Service's electronic
          newsletter CITE.  It would receive a bibliographic record but no
          holdings record since it would not be retained.  

     b.   State publications in the form of e-journals should be given
          second priority with regard to a bibliographic record plus
          archiving.  Examples of this type of electronic journal are
          Journal of Extension, Cooperative Extension System, University of
          Wisconsin and Social Sciences in Forestry, Forestry Library,
          University of Minnesota.  Since the publishers of these e-journals
          plan to archive them and these files are available via the
          Internet, the Committee was undecided whether NAL should also use
          computer space to archive these publications.  It was finally
          decided, however, that patrons would be better served by having
          the files locally available through a user friendly interface to
          an NAL computer.  This may be impossible to accomplish given
          funding, space and staffing considerations.


2.   Bibliographic record only, giving access information.

     a.   Electronic newsletters, whether they are published/sponsored by
          USDA, Cooperative Extension Service or any other organization
          should receive only a bibliographic record, keeping in mind that
          occasionally "newsletter" in the title of a publication proves to
          be a misnomer.


The e-journals mentioned so far in this section have all been free of charge. 
A subscription fee should not be a factor with regard to selecting an e-
journal if it is within NAL's scope, but it may be necessary to assign titles
with subscription fees a bibliographic record which includes remote access
information, rather than pay for archiving privileges.  In any case, such
archiving may not be permitted due to copyright restrictions as they are now
being interpreted by commercial publishers.  According to OCLC, libraries are
prohibited from mounting files of the Online Journal of Current Clinical
Trials on their own computers.  They conceded that it would be impossible at
this time to actually prevent a library or other organization from doing so
but they would consider it illegal and the downloaded files would not be
considered "official copies of Online Journal of Current Clinical Trials". 

When an e-journal is published in more than one format, the format should be
chosen which is most advantageous for NAL.   An example would be Laboratory
Primate Newsletter which NAL already receives in print format but is not yet
indexed.  The electronic version of this title does not offer graphics at all,
regardless of additional software or hardware,  so a decision would have to be
made as to whether the graphics were important enough to continue keeping the
printed version on the shelf.  The decision would have to take into account
the cost of shelf space and binding versus cost of computer space and
attendant maintenance.

This illustrates three points:  1. Each title must be considered as a separate
case.  2. Methods of comparing these types of costs will have to be developed. 
3. Liaison among branches of NAL will be very important with regard to
processing of electronic media.

If an e-journal offered in both print and online formats has not only a
subscription or licensing fee attached to its access but also an additional
fee for downloading, then print may be the only financial alternative. 
Examples of electronic journals in this category are those offered by Elsevier
through the TULIP Program and Springer-Verlag's Red Sage Project, the former
consisting of journals in materials science and engineering while the latter
offers journals in molecular biology and radiology.  It is also possible to
mount the journals in the TULIP Program on a local library computer for a
subscription fee, thereby making it necessary to calculate whether this costs
less than purchasing the printed version.  These represent a category of
electronic journal that not everyone considers truly electronic since they are
electronic versions of journals which are still published in print format
rather than journals which have been electronic since their inception.  

It should be noted that while, for purposes of  this Initiative, and in much
of the literature, electronic journals are defined as online publications,
some sources also consider journals on diskette and CD-ROM electronic,
pointing out the fact that terminology and definitions in this field are still
evolving.


Cataloging

Ideally ISIS patron interface for e-journals housed at NAL would simplify
access for patrons.  The "NEXT Workstation" should be able to access the
information.  The ISIS record could direct patrons to the NEXT workstation for
access.  For remote access the ISIS bibliographic and holding records could
provide information on how to access the material.  

Until a computer system is available to house full-text e-journals Acq/Ser
could download them to a file housed in a designated section on Cliff.

Acq/Ser will verify and confirm access to journals acquired by NAL as well as
remote journal titles.  They will also create a temporary record (using an "in
processing e-journal temporary call number" i.e. IPEJ No. 345) in ISIS which
would provide access information to the location of the material.  Remote e-
journals may receive only a "REMOTE JOURNAL" type call number.  At that point
Acq/Ser would send a message to Cataloging via electronic mail providing,
title, temp call number, and access information.  An e-mail processing form
(shell) could be developed to transfer this information to Indexing and
Cataloging, and after the item is cataloged the form could be updated and
transferred to DDSB, Reference, Information Centers, etc.

The cataloger, using a split-screen work station,  would access the material
and catalog it on-line on OCLC,  updating call numbers as needed.  The
cataloger or support staff will produce the record and it will go into ISIS
and AGRICOLA.

On a case-by-case basis, Acq/Ser will provide cataloging with separate
temporary records for both the printed and the electronic formatted journal. 
When information is known about other format options that information will be
included on the record in accordance with current CONSER practices.

Acq/Ser will confirm and process title changes and changes in mode of access
data to the Cataloging Branch via e-mail and will continue to note the change
on the ISIS record.

Updating information will be input into ISIS.  NAL will follow the CONSER
requirements in relation to input into OCLC.  Currently AGRICOLA cannot be
updated.  

The same call number will not be used as the cutter will be slightly
different.  According to current CONSER requirements, e-journals would receive
a separate cataloging record from the print version.  Cataloging will give e-
journals an "EJOURNAL No. __ " item level call number.  This will provide dual
access for Reference and other users who want to browse the e-journals. 
          
Cataloging will provide the same level of cataloging for USDA as well as non-
USDA e-journals.  Full CONSER level records will be provided as appropriate in
OCLC, ISIS and AGRICOLA.

When a new computer system is available and/or when enough materials are
acquired for the collection to determine the nature, scope, quality access and
preservation requirements,  all short term policies and practices will be
evaluated and redefined as needed.


Indexing

When NAL receives a journal in both print and electronic version, Indexing
will choose to work with the print version until staffing, technology and
funding make choosing the electronic version preferable.  Indexing recommends
that all e-journals chosen for indexing be archived indefinitely at NAL rather
than simply providing remote access information in the bibliographic record. 
The ability to receive graphics is essential to the indexing process. 
Therefore any software that is required with individual e-journals will be
absolutely necessary.   The Windows application developed for linking ISIS
records and scanned abstracts with the full text of the e-journals will be
investigated.


Reference

The Software Demonstration Center should be changed into an Electronic Media
Center.  Implications for Reference will need to be studied and more staff may
be required whether it is staffed by Reference or is a unit independent from
Reference.  A vending machine for floppy diskettes should be installed.  The
possibility of charging for paper copying should be evaluated and if  found to
be appropriate, a mechanism should be identified and implemented.


Document Delivery Services

DDSB staff can access electronic journals by using several individual Internet
accounts.  When an ILL request for an e-journal is received DDSB staff will
refer the requester to the site where the journal is archived.  USDA requests
for articles from electronic journals could be filled by having DDSB staff
access the archive where the article is stored, and either printing it out for
the patron, downloading it on a floppy disk, or transferring it electronically
to an electronic mail account for the requester.

NAL patrons have a right to expect that,  despite the novelty of electronic
media, they will receive requested documents in at least the timeframes to
which they have been accustomed for traditional print documents.  In order to
process document delivery requests for electronic media in a timely fashion,
DDSB staff and contractors need to have direct access to all publications in
the collection.  At the present time, DDSB staff can absorb the extra work
without a contract modification.  The materials in the SDC should be moved to
the general collection, in an area to be determined appropriate for an
electronic text collection.  Frequently used material may be stored in the
SDC, and "retired" to the general collection when its popularity wanes.  This
would be equivalent to the concept of the Periodical Reading Room.  NAL may
choose to acquire two copies of very popular items, and have one copy for SDC
and a collection copy for document delivery.

All DDSB supervisors are now using the Cliff electronic mail system. 
Additional training is needed in Internet resources, such as gopher, WAIS,
Wide World Web; and the use of file transfer protocol (ftp) and telnet. 
Training is needed in the short term for the supervisors and technicians in
DDSB.  As demand for electronic media continues all DDSB and contract staff
will need training, both in Internet resources, and basic computer skills. 
For the long-term, NAL needs to commit itself to keeping its  staff up-to-date
on developments in the publishing industry, on the Internet, and on the
information needs of its scientists.  This requires budgeting for training,
professional meetings, and travel.

Until some standardization of publishing electronic media is achieved, each
new e-journal received by NAL may require training to use.  The current
organizational structure of NAL, and the resulting workflow, may cause each
branch to make separate and duplicative efforts to learn to use each new
acquisition so that they can perform their processing tasks, or to respond to
reference or document delivery requests.  NAL needs to assess the best way to
handle such training, and perhaps assign an individual or section to be
responsible for it.

NAL pays copyright fees for copies it requests from other libraries through
interlibrary borrowing (ILB).  If NAL has an active subscription for an
electronic journal, there should be no costs from copyright fees other than
the subscription or lease price.  DDSB has not yet had to order an electronic
journal on ILB.  In the future, it is expected that such requests will become
a gradually increasing part of the ILB workflow.  However, depending on how
Internet access grows among USDA employees, the possibility exists that users
may access these electronic journals themselves, and pay for these costs, if
any,  directly.  The development of commercial publishing needs to be watched
closely during the next few years to predict what these costs will be, and who
will be paying for them.

As the volume of document delivery requests for electronic media increases, an
amendment to the DDSB lending services contract will need to be written.  In
order to estimate the additional cost to the contract, DDSB will need to know
the estimated volume of requests, the amount of time spent processing
requests, and the level of skill required to perform the tasks.  It is not
certain what the estimated volume of requests per year will be for this
activity.  Also, it is not known if the proliferation of Internet accounts
will mean that the patrons will be able to access this material themselves,
without the assistance of DDSB or any NAL intermediary.  Access to the NAL
building network by ARS libraries or other outside groups may also make
intermediate assistance unnecessary.  In the short term, while DDSB staff are
performing these tasks, DDSB supervisors will need to track the time needed to
fill these requests, and the level of skill required to perform these tasks.


Storage and Systems 

The HP has a lot of unused capacity and whether NAL wants to use that capacity
for e-journal storage will have to be decided.  Issues to be addressed
include:

     -    no search software for full text on the HP
     -    no link from/to ISIS system for full text searching
     -    no staff to operate this type of system

ALF is a bulletin board system (BBS).  BBS's are designed for simple
(2400-bps) dial access and short file transfers.  ALF may not be the best
system for delivery of this volume of data.  The BBS software may be
overwhelmed by a full-text search and we would probably need to upgrade the
hardware considerably as well as add staff to manage the increased usage. 

A UNIX based system (with proper power protection and data back-up) would be a
much better choice for on-line access because the hardware and operating
system are designed for multi-user access of large files.  Dial access issues
can be addressed by using faster communications hardware than is common in the
PC world.  The issues of full-text search software and staff support remain a
problem.  

A custom interface or full-text search engine will be necessary.  We will also
need full-time staff support for such an operation.  Limitations will be
related to the choice of hardware/software.  UNIX is a multi-user,
multi-tasking operating system that will allow access, but the choice of
hardware will determine performance limits.  

All the training currently provided by ISD is aimed at staff improvement. 
Documents aimed at helping patrons will have to be created.  Currently, staff
at both the Reference Desk and Circulation/Information Desk provide patron
assistance on using the card catalog/ISIS.  One hopes that any online system
will be as easy to use as ISIS.  In-house delivery will be a relatively simple
matter from any workstation on the backbone.



Practical Experimentation

The following e-journals should be used for initial experimentation and
documentation:

1.   Extension Review - USDA Extension Service, Washington, DC

     *    No longer available in printed format.  Last printed issue v.63,
          no.1, May, 1992.  The first and only electronic issue is a copy of
          the last printed issue.

     *    Articles not authored though contributors are listed.

     *    No OCLC record of electronic version. (As of March, 1993)

     *    Future uncertain due in part to the reorganization of USDA and the
          Extension Service.  At this point a low priority.  (Communication
          with CIT, May, 1993).

     *    There are at present no plans to archive this title within the
          Extension Service.

     *    Access: almanac@ecn.purdue.edu  Command: send ermay92


2.   Journal of Extension - Cooperative Extension System, University of
     Wisconsin, Madison, WI.

     *    Continues to appear in print.  NAL receives print copy free of
          charge via Gift & Exchange Unit.

     *    First electronic issue Winter 1992, first full issue Spring 1993. 
          Five issues in all will be published during the pilot study.

     *    No OCLC record for electronic version as of March, 1993.

     *    Authored articles.

     *    Upon subscription, a table of contents message arrives
          automatically.  Subscribers then may request via e-mail full text
          of one or more articles. 

     *    Last five years are archived and indexed.
 
     *    Copyright statement says, in part, that single copies may be
          reproduced in print or electronic form for educational or training
          activities but inclusion of articles in other publications,
          electronic sources or systematic large-scale distribution may be
          done only with prior written permission.

     *    Access: almanac@joe.uwex.edu  Command: subscribe joe.


3.   Social Sciences in Forestry: a Current Selected Bibliography and Index.
     - Forestry Library, University of Minnesota, St. Paul, MN

     *    Last printed issue December, 1991.

     *    Quarterly.  Archived since no. 1, 1985.

     *    No OCLC record for electronic version as of March, 1993.

     *    No discernable copyright statement.  

     *    Rather than being available via an e-mail subscription, SSIF is
          based on a gopher distributed information delivery system which
          allows full text searching of the annotated citations.

     *    Access: telnet consultant.micro.umn.edu or telnet info.umd.edu 
          (University of Maryland's gopher).



4.   Laboratory Primate Newsletter - Brown University, Providence, RI

     *    Also available in print.  NAL receives print copy free of charge
          via Gift & Exchange Unit.

     *    Electronic version does not include graphics.

     *    Quarterly.  

     *    Authored articles.

     *    Archived since January, 1991.

     *    No OCLC record for electronic version as of March, 1993.

     *    Upon subscription, a table of contents will arrive automatically
          each quarter, with the capability of then requesting individual
          articles.

     *    Copyright statement: "All material is copyright, but may be
          circulated electronically, or printed out and reproduced for
          classroom and other non-commercial purposes, as long as reference
          is made to the source".

     *    Access: listserv@brownvm.brown.edu  Command: subscribe LPN-L Your
          Name. 


     Laboratory Primate Newsletter was selected as an example of an e-journal
     which is within NAL's scope yet is neither a 1USDA nor a state
     publication.

At present, there appears to be no e-journal published by a commercial
publisher which concerns agricultural subjects such as crop science, animal
husbandry etc.  E-journals such as those published by OCLC, Online Journal of
Current Clinical Trials and the latest venture, Online Journal of Knowledge
Synthesis for Nursing, are not within NAL's scope and therefore, NAL should
not subscribe to them.



                         Members of the Committee

                             Rae Dubois, Chair

                                John Forbes
                               Janet Berkson
                               Julie Mangin
                               Daniel Starr
                             Veronica Lefebvre





                           CHAPTER 4  -  CD-ROM

Original texts and reproductions of print documents disseminated on CD-ROM

Summary

This Committee was charged with developing policy proposals and procedures for
the acquisition, processing, and dissemination of publications on CD-ROM. The
group examined a wide variety of issues and made extensive headway toward
integrating this medium with other formats in the Library.  One of the major
difficulties with this type of media is the complexity of procedures required
to process these materials.  While the committee made substantial progress,
more time is clearly needed to implement, expand and modify proposals set out
in the three reports prepared by the Committee. 


Recommendations

1.   Develop a policy statement outlining processing (Acquisitions & Serials,
     Cataloging, Indexing) priorities which specifically address media types. 
     If processing priorities are changed, then staff can be shifted to work
     on materials deemed most valuable to the mission of the Library.  This
     is particularly important in a period when increases in staffing are
     unlikely. 

2.   Fill in the details of the workflow which has been sketched out as a
     skeleton framework for the processing of these materials by the
     Acquisitions/Serials, Cataloging and Indexing Branches.  Cataloging and
     indexing responsibilities would be determined based on how earlier print
     publications issued by a publisher were processed in the Library. 
     Responsibility for processing a disk could include having some documents
     cataloged, some indexed, and some items checked-in as serials.

3.   Begin processing some CD-ROM publications through Cataloging. It is
     assumed that eventually publications on CD-ROM will have the same
     processing priorities assigned as other formats. If this policy is
     adopted, other materials will have to be assigned a lower processing
     priority.

4.   Establish the impact of CD-ROM publications on the Gift and Exchange
     process.  A major problem will be how to measure the value of a CD-ROM
     against print publications.

5.   Establish an electronic media line item in the acquisitions budget.  In
     FY 93 $65,000 was spent  for CD-ROMs.  For the Library the size of NAL,
     it would not be unreasonable to spend $175,000 on electronic
     publications.

6.   Update the collection development policy for electronic media.

7.   Establish the cost of check-ing in serial titles on CD-ROM publications
     by performing experiments using the CGIAR CD-ROM.  Check-in of these
     publications will be valuable and should be done on a regular basis. 

8.   Establish electronic versions of print publications as an important
     component of the Library's preservation and space policies.  Process
     electronic versions of NAL-owned print publications as added copies on
     ISIS.  Each added copy item record will have to be individually created
     by Cataloging Branch staff.  Added copy indexing records should be
     created through computer generated scope changes on ISIS. 

9.   Establish the actual cost of indexing an electronic publication by
     performing time/cost experiments with the CGIAR CD-ROM.   Valuable
     citations will be lost if in-scope electronic publications are not
     indexed.

10.  Assign the Software Demonstration Center staff the responsibility of
     learning CD-ROM programs, and instructing units within the Library how
     to use them.  This more proactive role would save the cost of each unit
     having to learn how to use every new software package the Library
     purchases. 

11.  House the electronic media collection (including all software) in a more
     accessible and environmentally protective area.  It is the committee's
     feeling that the SDC should not be the repository of NAL's software
     collection, but if needed, it should be a demonstration center for
     software.

12.  Test the process of copying articles from CD-ROMs for the purpose of
     document delivery.  It will initially be a very expensive process to
     learn, and will be done by DDSB staff.  If the volume of requests for
     electronic media increases, the contract will have to be amended to
     permit the contractor to search CD-ROMs and copy files.

13.  Training in DOS, Windows, and basic computer skills needs to be an
     ongoing activity.   There will also be substantial training costs within
     each Branch.  Staff need to attend conferences in order to keep abreast
     of changes in the fields of computer technology and the applications of
     these technologies in libraries.

14.  Evaluate whether Reference staff should maintain most of the policies
     that are in place at present,  and whether SDC staff will be responsible
     for working with patrons who need to use electronic materials other than
     the standard reference tools.

15.  Obtain the hardware necessary to accomplish these recommendations. 

16.  NAL managers must emphasize to staff the importance of the role it
     perceives electronic publications will play in meeting the Library's
     mission.

17.  NAL should effectively advertise to its public the electronic
     information services it can deliver.

18.  Assign responsibility for staying abreast of technological developments
     within the computer industry to specific individuals who are also
     responsible for sharing that information with ALL NAL STAFF.


Current Situation

It is essential to have written policies on the acquisition, cataloging,
indexing, storage, public use, lending, copyright, and preservation aspects of
CD-ROM with publications. No medium that NAL has thus far received will
require as much coordination among units as will this one.

Currently, NAL acquires a variety of CD-ROM products.  Those published
serially are checked-in on ISIS from eye-legible printed information on the
disk. NAL spent approximately $65,000 in FY93 on CD-ROMs.  To facilitate
check-in, a dual page monitor, CD-ROM reader, and accompanying software were
purchased.  Approximately 4 hours per week are spent ordering, checking-in,
claiming, recording invoices, and otherwise following up on these products.

No publishing standard exists for these types of CD-ROMs due to the fact that
this is a new publishing medium and many publishers are utilizing this format
in different ways.  Much material is "dumped" into CDs with little thought
regarding standardization of searching and retrieval strategies. 
Consequently, there is a considerable cost in professional staff time spent
figuring out how to use these products, and once that is done, training staff
as needed. Time spent on acquisition and check-in activities will increase
enormously as we continue to receive large numbers of CDs which require
analysis of content, such as the FAIRS and CGIAR disks.  Check-in of the 
content of these unusual CDs is a complex task and entails the use of time
over and beyond that which is normally used for check-in of paper
publications.

If the materials budget remains static, which may be the net result of
foreseeable budget cuts,  choices will have to be made between print and
electronic (CD-ROM) titles. There will be price increases for the CDs
presently purchased, and there will be increased pressure from the Public
Services Division and patrons to purchase new titles as they become available. 
There is also the real possibility that CD-ROM networking will be started here
on a small scale, and this will invite increased costs as CDs are purchased
for networking purposes.

It is very important that the serial contents of these CD-ROMs be checked-in.
At a minimum, selected titles should be checked-in on a regular basis.  More
work should be done to determine which titles should be checked-in and whether
retrospective serial titles should be checked-in.

Currently, DDSB staff do not have direct access to CD-ROMs in the NAL
collection. They are stored in the Software Demonstration Center, and SDC
staff must retrieve items for use. Items are then charged out to DDSB.   NAL
patrons have a right to expect that,  despite the novelty of electronic media,
they will receive requested documents in at least the timeframes to which they
are accustomed for traditional print documents. In order to process document
delivery requests for electronic media in a timely fashion, DDSB staff and
contractors need to have direct access to all publications in the collection.

As the volume of document delivery requests for electronic media increases, an
amendment to the DDSB lending services contract will need to be written. In
order to estimate the additional cost to the contract, DDSB will need to know
the estimated volume of requests, the amount of time spent processing
requests, and the level of skill required to perform these tasks. The
complexity of CD-ROM searching strategies may affect whether these materials
are included in the contract. Each CD-ROM title that is received comes with
its own software and the method or retrieving documents must be learned for
each title.  According to a Cataloging Branch estimate, it may take an average
of 4 hours to learn how to use each new CD-ROM title acquired by the Library. 
For this reason, assistance from SDC staff in learning how to use different
software packages would be helpful.

It is not clear if the proliferation of Internet accounts will mean that
patrons will be able to access this material themselves, without the
assistance of DDSB or any NAL intermediary.  Access to the NAL building
network by ARS libraries or other groups may also make intermediate assistance
unnecessary.  In the short term, while DDSB staff are performing these tasks,
DDSB supervisors will need to track the time needed to fill these requests,
and the level of skill required to perform these tasks.

The inability of the Software Demonstration Center (SDC) to be able to load
and display computer software will be an ongoing source of frustration to
users unless its hardware is upgraded.  This Committee lost a significant
amount of time waiting for the SDC to make available a computer with enough
memory to load the CGIAR (Consultative Group on International Agricultural
Research) software.  After the software was finally loaded onto an IBM XT, the
committee realized that other computer hardware would have to be found because
of the XTs slow response time.  The end result was that a 386 in the
Information Systems Division was used, and a delay of almost three weeks
passed before work could be done analyzing the CGIAR disks.


Issues

Selection/Acquisitions

Criteria for selecting CD-ROMs should reflect NAL's collection development
policy concerning subject coverage and level of collecting intensity. Taking
into consideration the price of many of the CDs, however, it is important to
realize that budget constraints may limit the "comprehensive" nature of the
CD-ROM collection, just as they limit acquisitions of print publications. It
is not known at this time what volume of information will be published on CD-
ROM, or whether this format will become the publishers' format of choice.
Currently, NAL purchases many indexes published on CD-ROM, because searching
CD-ROM is far easier than searching hard copy indexes, and also because the
purchase of a CD may be cheaper than searching online. But the information
which is cited in the indexes is usually available at NAL in hard copy format.


NAL's selection policy for machine readable materials issued in 1985 needs to
be updated or clarified, especially sections A.5 and B.1.


CD-ROM Purchases

Although a CD-ROM may have original content, and/or be a full text document,
it is often times also issued in paper and microform formats, and may even be
available online. The decision concerning which titles to purchase and which
format(s) to obtain must be made on a case by case basis. There are several
considerations:

--   CD-ROMs (generally) are very expensive.

--   CD-ROMs may be offered at a discount price if you also purchase the hard
     copy. How much discount? Do you want to pay the combined price?  The CD-
     ROM by itself may be cheaper than the combined price.

--   The purchase of a CD-ROM may not be a purchase at all, but only a lease.
     Depending upon the title, the entire run of a serial, or a portion of
     back issues, may be forfeited if a renewal is not made. So payment is
     made to keep current, but the collection does not really belong to NAL. 
     For some titles, the back issues are owned, but the current run must be
     returned. There are also titles which do not have to be returned at all.

--   How often is the information on the CD-ROM updated? How often is it
     cumulated?  Is the CD-ROM a good alternative to an online resource if
     up-to-the-minute information is often needed? What are the
     costs/benefits of paying for online searches versus the cost of a CD-
     ROM? 

--   How many copies are needed? Would keeping alternative formats satisfy
     the need for multiple copies? What role does the purchase of the CD-ROM
     play in the Library's overall preservation policy? Would multiple copies
     of the CD-ROM itself, on a network, or on a standalone basis, best
     satisfy patrons and NAL staff?

--   The cost of CD-ROM readers, printers, workstations can be considerable.

 

Check-in 

A check-in record is a record of receipt, consequently, a record of holdings,
of publications which are published serially, or in more than one physical
part. It informs patrons what they can expect to find at NAL, and reminds NAL
personnel to "claim" or order missing parts. (However, it does not account for
parts which become missing, for whatever reason, after they are received). 
Taken in this vein, the Serials Unit at NAL does, and will, check-in CD-ROM
titles which are serials, works-in-parts or monographic series. This does not
require special equipment, as the designation of the part of the title that
the physical CD-ROM contains, is on the CD-ROM's label.

Some CD-ROMs contain collections, or runs of serial titles, or parts of serial
titles, or various combinations of the above, and it is certainly useful for
the librarian and the patron to know what is contained therein. These CD-ROMs
may contain runs of titles which are held by NAL in hard copy, not held by NAL
but desired by NAL, or missing issues of titles held by NAL that are now out-
of-print.

Examination of the contents of a CD-ROM, and serial check-in of the CD-ROM's
contents is not a function as we presently know it.  Rather, it is
descriptive, and is more in the realm of cataloging or indexing.  This may
not, of course be the end of it.   As this form of publication develops, we
will be forced by circumstances to modify our organization and technical
application of what is required.  ISIS may contain several class numbers that
it presently does not have, which describe contents and link holdings from one
format to the next, and the Serials Unit could modify its procedures.  Check-
in staff will need CD-ROM readers in close proximity to their work stations,
and the configuration could have dual page monitors as part of the NAL online
and ISIS network. It is too early to require that CD-ROM content be reviewed
on a CD-ROM reader and then be checked-in, but we do know that many runs of
serials are now on CD-ROMs, so some kind of accommodation must be made to this
concept.


Licensing Agreements

Fair use clauses in the copyright law may enable NAL and other libraries to
use information on a CD-ROM as they would information in a book or journal.
However, licensing agreements, not to be confused with copyright, are often
more constraining. It would appear publishers are using licensing agreements
to undermine and circumvent the copyright law.  Issues which need to be
addressed are:  

--   Do we give up copyright if a more stringent license agreement is signed?

--   Who at NAL should sign licensing agreements?  What liability does that
     person have? 

--   To what extent can NAL modify license agreements?

The library community is quite concerned with the phenomena of publisher
control and ownership of information. The Library of Congress is developing
its own generic license for CD-ROMs which it will present to publishers. NAL
should get permission to adapt this agreement and use it.   


Gift and Exchange

At this time, the effect of the CD-ROM phenomenon on the Gift and Exchange
Program at NAL is unclear.  As CD-ROM titles become known to the G&E Unit, G&E
will request appropriate titles as gifts. Once a CD-ROM has been published, it
is inexpensive to duplicate. However, publishers are unlikely to give NAL a
CD-ROM with an expensive purchase price.

Presently, NAL receives a number of government documents on CD-ROM through the
GPO depository program. This is usually the only format available, with the
print format having been abandoned in favor of CD-ROM. Care must be taken
however,  to obtain all CD-ROMs published by GPO that are in scope for NAL as
CD-ROMs are not always considered publications.

Departmental Regulation 1020-1 dated March 23, 1990, specifies that two copies
of all substantive USDA publications must be forwarded to NAL. This directive
has to be modified to specify that CD-ROMs (as well as other electronic
formats) are publications.  There must be no doubt as to the importance of
this information, and USDA  agencies must be reminded to forward their
electronic publications to NAL.   It is necessary to put such a policy in
place for land-grant experiment station and extension service publications in
electronic format. Formal lines of communication should be put in place so
that agencies and universities can be contacted easily, they realize the
importance of their CD-ROMs to NAL, and remember to send them to the Library. 

In the past, USDA agencies have voluntarily mailed their own publications to
exchange partners. With some agencies no longer willing to incur these
expenses, NAL will have to play a larger role in the exchange process in the
future.

Since some USDA agencies publish only on CD-ROM, there may be instances where
there can be no basis for an exchange relationship because the foreign country
lacks the appropriate computer hardware.

One question raised is what is the exchange value of a CD-ROM containing
individual publications issued by an agency in the U.S. or abroad?  What about
the case of an agency that issues in CD-ROM format a retrospective of all
documents it has issued?




Processing Responsibilities

The decision on whether items should be cataloged or indexed will be assigned
to an interdepartmental review committee, consisting of representatives from
the Indexing Branch and the Cataloging Branch, and perhaps two representatives
from the Public Services Division.

New publications issued for the first time on CD-ROM will be evaluated for
cataloging or indexing based on the history of the types of materials issued
by the publisher, how these publications have been processed in the past at
NAL, whether the publications on the disk are indexed by other agencies, and
available staff resources to bibliographically process the disk.  The more
difficult publications to evaluate will be those which have not been
previously issued in any format.

The Indexing Branch generally bases its decision to index publications on
whether they fit within pre-established criteria. They must be within the
AGRICOLA scope, have format/structure, have acceptable publication status, fit
pre-established worldwide responsibilities, and they must be "doable," taking
into account staff and equipment limitations. NAL has committed to index items
produced by land-grant universities and their affiliated extension services
and experiment stations.. Currently they are received in paper-only format,
but in the future they may be received in electronic-only format, e.g., FAIRS
CD-ROM.

The Cataloging Branch has no such guidelines for determining whether a
publication should be cataloged. Processing priorities need to be reevaluated
as they relate to the mission of the Library. Beyond reviewing processing
priorities, the following questions need to be addressed. In what instances
should publications on a CD-ROM disk not be cataloged or indexed? What
responsibility does NAL have to provide cataloging records to OCLC, AGRICOLA
and other bibliographic databases? Should NAL continue to catalog all
agriculture related publications from around the world?

In many instances an entire disk may be processed either by the Cataloging or
Indexing Branch.  In other cases, responsibilities will be mixed, and
responsibilities of units will need to be formalized so that an entire disk
can be bibliographically processed.

Once a decision is made to catalog and/or index a disk, the Acquisitions &
Serials Branch will route subsequent disks of the same title to the
appropriate location.  Procedures on how CD-ROM disks are physically processed
need to be developed, so that both ACQ/SER and Cataloging adhere to the same
guidelines.  A task force should be named for this purpose.

Depending on preservation policies established by the Library, and the
determination that these activities fit within the goals of the Library,
Cataloging Branch staff can/will create added copy records for electronic
publications in a cost efficient manner that will be linked to previously
published paper copy bibliographic records.  Indexing Branch records will have
added copy records created for the electronic version of a previously issued
paper publication through batch computer techniques yet to be used by NAL.
This will be accomplished by scope changes made to sets of indexing records.
The result of these bibliographic maintenance activities in both the
Cataloging and Indexing Branches will be to allow weeding of paper copies from
the stacks in the future. Procedures for added copies will need to be further
defined.

Discussion has begun on the possibility of creating a new class in ISIS where
indexes of disks the Library chooses neither to catalog or index can be loaded
("dumped").  This will allow patron access to these materials. The time frame
and technology required for this new class is at present unknown.


Indexing

Policies for indexing CD-ROM publications were non-existent until electronic
media issues began being explored throughout the Library. Current indexing
policies are based on indexing the printed copy.  Whether this continues to be
acceptable depends on the outcome of this Initiative.

A goal for publishers of agriculture related CD-ROMs should be to standardize
their presentation of bibliographic information within a document.
Publications within the disk should contain easily identifiable bibliographic
data such as: identification number or file name, item title, authorship,
authorship affiliation (optional), source information,  publisher, location of
publisher, publishing date, series statement, and whether its also available
in paper. Edition information should also be made available; i.e., is the
publication a new edition, revision or reprinting?  A future goal is to be
able to transfer electronically the coded data on the disk to a MARC
bibliographic record.

Publishers should format documents so that the title of the item is at the
beginning of the publication.  If the publication is in a foreign language, an
English translated title should be provided. All authors should be listed
completely and accurately, preferably in the vicinity of the title. Currently,
it is Indexing Branch policy to provide author affiliation for all records
processed for AGRICOLA. The author affiliation should be clearly stated and
located near its appropriate author's name.

Publishers should also provide a detailed table of contents (menu) on each
disk.  Contents should be sorted by "newly published," "revised, or
"reprinted."  This will help eliminate duplication of indexing records, by the
ability to focus only on the new and revised publications. 

Before items are indexed they are checked for call number, priority status,
and routing to indexers. Items selected for indexing must be attached to the
appropriate host cataloging record.  If the items are only available
electronically, then a procedure has to be established for notifying the
appropriate indexer of items awaiting indexing.  Some disks will contain
publications in a variety of subjects requiring indexing by several indexers. 


Once the decision has been made to index items on a particular CD-ROM, routing
within the Indexing Branch must be established. Disks will need to be
transferred electronically, either through a LAN, ISIS or some other
electronic system. This is the most efficient way to process CD-ROMs.
Otherwise, multiple copies of specialized hardware and software will need to
purchased in each unit that processes a CD-ROM, i.e. Serials, Cataloging,
Indexing, Document Delivery, Reference, etc.

Problems arise if the on-screen version is different from the printed copy.
Procedures need to be established to indicate format variations. If the CD-ROM
has an exact duplicate printed version, then indexing would be done from the
paper copy.

Currently, there are two levels of indexing --- minimal and full level.
Criteria have been established based on format, source, scope, etc. These same
criteria would be used in determining the indexing level for CD-ROM items.

The actual indexing of the item also poses problems. Currently, indexers
produce written indexing records which in turn are keyed by technicians.  In
the very near future, the capability for online indexing into the ISIS system
will be in place. If the CD-ROM is available online, then an interface needs
to be established between the ISIS system and CD-ROM search software. Also, a
split screen monitor and associated software is required so that the indexer
can look at the CD-ROM product and enter the indexing record into ISIS using
only one computer. 

If the CD-ROM must be mounted on a local computer in the Indexing Branch, then
additional hardware/software will need to be purchased to support this local
loading. Additional space will be needed to house this additional equipment.
Adequate employee training will need to be provided. The most efficient method
of indexing this material will need to be determined.  

Quality control needs to be maintained for indexing records. Currently, the
MARC format is used to produce AGRICOLA records. CD-ROM publishers should
follow established printed publishing guidelines. A definite format should be
established for the CD-ROM and that format should be followed consistently. 

Currently, the Indexing Branch scans author provided abstracts from over 300
journals and adds them to AGRICOLA. Mechanisms need to be defined as to what
procedure will be followed if these journals are only published in CD-ROM
format. Additional hardware/software might be needed to download and upload
these abstracts. 

Graphics play a key role in the indexing process. Pictures, charts, graphs,
tables and illustrations often provide information that is not available in
the text. Full-page monitors will be needed in order to view these graphs.

Once the CD-ROM has been indexed and indexing records have been processed into
ISIS, procedures need to be established for the routing of the disk. If the
disk was physically in the Indexing Branch, then its next location needs to be
established, i.e. Electronic Media Center, Document Delivery Services Branch,
Reference, ISD, etc.  If the disk was loaded on a computer serving a LAN, then
appropriate personnel need to be contacted to indicate that work on the disk
has been completed and it no longer needs to be on a LAN. Procedures regarding
deletion, transfer to other files, etc. need to be established. 

Currently, there is no mechanism in place for noting multiple format copies,
i.e., paper and electronic formats.  This information will be added to
retrospective indexing records by utilizing efficient, batch processing
computer techniques. To manually process retrospective indexing records would
be a monumental task.


Cataloging

The policies listed below are generally adapted from existing Cataloging
Branch guidelines for processing monographs and computer software. While the
format of these publications is new, Cataloging Branch staff have been
cataloging the print version of these publications for years.  

Electronic media materials should be assigned the same cataloging priority
treatment as print materials, based on content, scope, etc.  Minimal and full-
level cataloging policies will be determined irrespective of format. 
Publications on CD-ROM will be cataloged by appropriate staff based on the
level of cataloging determined for the documents on the disk.  Analytics of
series traditionally cataloged by NAL support staff will continue to be
cataloged by them.  Cataloging Branch staff will create bibliographic records
for both the collective title of a disk,  and individual titles on a disk as
necessary.

Electronic publications requiring full-level original cataloging will be
cataloged by professional staff. Chapters 9 and 13 of the Anglo- American
Cataloging Rules will be used in preparing records for newly published CD-ROM
titles. Publication information in particular, and other bibliographic data
may prove difficult to identify in electronic formats, where this was a
relatively simple task for print materials. This problem will need further
exploration, and may be handled in different ways depending on the display of
the data.

Cataloging Branch staff will catalog the paper version as the original if the
Library receives both copies. The Branch may choose not to follow AACR2/MARBI
guidelines for multiple versions.  If a paper and electronic version of a
publication are similar, but do not meet AACR2 criteria for an added copy,
Cataloging Branch staff may choose to locally treat these similar items as
added copies to save the cost of creating new cataloging records.

A particular concern at NAL may be the loss of bibliographic series which have
traditionally appeared on print versions, but may be difficult to locate on
the electronic version.   Adequate computer equipment and staffing will be
required to carry out these tasks.  The amount of cataloging of CD-ROMs should
be adapted to staffing levels.  Cataloging Branch staff will use split-screen
monitors to create MARC bibliographic records while simultaneously viewing the
CD-ROM publications.  Staff training will be an important element in using
this new technology.  Staff should use constant data records in creating new
bibliographic records when possible.  Processing a disk which appears on a LAN
would require similar processing procedures.

CD-ROM products will be assigned a Library of Congress classification number
based on their subject content. Accompanying guides, manuals, etc. will be
housed with the disk in the Electronic Media Center.  An unresolved question
is whether the number of the disk on which the publication is stored should be
included in the call number.  With CD-ROM sets being updated constantly, the
disk number could change.
 
The amount of help the cataloging record should provide NAL patrons and staff
needs to be determined. Should the file name of the publication be displayed
in a note?  How much time will it take to provide this information?  On the
FAIRS disk, should the location of articles within a menu be noted.  For
example, an article on beneficial insects appears in the "Extension Handbooks"
menu.   Several members of the NAL staff are working on a means of encoding
bibliographically significant data and having it electronically transferred to
a MARC record.  

Finally, there is some question as to whether any of these CD-ROMs should be
cataloged.  Will the ability of computers to create searchable keyword indexes
negate the need to catalog the contents of CDs?


Copyright

It was suggested that NAL not circulate its only copy of a CD-ROM when it is
copyrighted.   If material from a copyrighted disk is given to a patron, can
it only be on a diskette which must be returned to the Library?  Can a higher
fee for copyrighted disks be assessed which will allow unlimited transmission
of a disk's contents?

A question raised which needs to be answered is whether an uncopyrighted
article/book on a copyrighted disk is protected by copyright?  Robert Oakley
states in his article "Copyright Issues for the Creators and Users of
Information in the Electronic Environment" that the present copyright law
should be applied to electronic media and only minor changes to the law are
necessary.

In compliance with the copyright law, NAL presently copies no more than 50
pages from a book or journal article  What guidelines should be used for
electronic publications?  How does one determine what percentage of an
electronic publication can be copied?

Electronic publications that are available on a LAN also create a copyright
problem.  Technically,  several individuals can copy the same publication. 
How is copyright infringement averted?  Does this mean that only uncopyrighted
publications can be on the LAN? 

Keeping current with changes in the copyright law is essential, especially in
light of licensing agreements. There may be a need to contact USDA lawyers for
rulings on how the copyright law should be applied by NAL to electronic media.



Reference

Reference Section staff currently work with patrons on only a select few disks
(23 titles) which are part of the Reference Collection.  Most of the CD-ROM
products in the Reference Collection are databases, or solely reference
materials.  All other CD-ROM products are stored in the Software Demonstration
Center. The Reference Branch has published a policy statement on the use of
Reference Branch CD-ROMs.

Reference Branch staff spend as much time with a patron as necessary in order
for the user to understand how the CD-ROM product works.  Cheatsheets, batch-
program screens and other aids are designed for end-user searching.  The need
for classes at one end of the spectrum, and hand-holding at the other end,
might merit additional study.

Some CD-ROM products are owned by the Information Centers, but the number is
unknown.   Most CD-ROMs not housed in the Reference Collection are stored in
the Software Demonstration Center.  SDC staff help patrons to use the many
specialized electronic media products held by the Center.   The DC Reference
Center owns a few CD-ROMs, which may be used by patrons. They provide
instruction to first-time users in the same manner as at NAL.


Document Delivery

DDSB considers having cataloging, indexing records and serial holdings records
within a disk critical. 

CD-ROMs housed in the SDC need to be made more accessible to DDSB staff and
patrons to fill requests. It would make more sense to transfer all or part of
the collection in the SDC to the general collection to some sort of electronic
text center. Perhaps only the software used for demonstration purposes would
be appropriate in the SDC. If electronic media were stored in an area to which
DDSB has direct access, DDSB staff or contractors would handle document
delivery of electronic media. It is expected that electronic media will be
more prevalent in the future, making space in the SDC inadequate for storage. 

The question of how much of a disk can be copied still needs to be decided. 
How are the CONTU guidelines which limit copying to 50 pages of a text applied
to electronic media?  If a CD-ROM is loaded on an network, will NAL loan it? 
What if the title is designated "reference" or "reserve"?  Can we use the "CD
maker" that ISD used to create the "Plant-It CD" to making circulating copies?

Concerns were raised about the ability of Document Delivery to have access to
disks being processed.  If disks remain in Technical Services for a long time
because of the large number of bibliographic and indexing records that need to
be created, this could impact on document delivery.  An NAL-wide network in
which all units would have access to a CD-ROM publication is preferred.
Perhaps loading these records into ISIS in some manner would have a similar
impact.  An NAL goal should be to provide access to in-process records,
whenever possible.  


Preservation

Library staff need to have a clear statement on the role that CD-ROMs play in
NAL's preservation policy.  Will paper issues of publications in the stacks be
removed from the collection and replaced with CD-ROMs?  Who will determine
what publication can be removed from the collection?  What will be the
magnitude of the replacement of paper publications with electronic media? 
What is the durability of the CD-ROM as opposed to print publications?  Is CD-
ROM simply an interim medium that will have to be converted to yet another
electronic format?   


Quality Control

One issue that has never been resolved, is who is responsible for quality
checking materials once they are received by the Library?  Who is responsible
for determining that the Library has received all the publications on a disk
that it has paid for?  Has the Library decided because of staffing limitations
that it will not quality check these materials? 


Standards

It is important for units within NAL to promote the development, evaluation
and implementation  of standards of materials published in electronic format.
An electronic publishing committee has been established to develop standards
for extension publications.  The impact these standards will have on the rest
of the agricultural publishing community is unclear. The USDA Extension
Service, land-grant agricultural publishers, USAIN and Agricultural
Communicators in Education should all be consulted in the development of a
standard publishing format.


Advertising

NAL should advertise its electronic media services.  It should only promote
services that can be delivered within the constraints of the Library's budget.


Computer Hardware, Networks

Critical to processing and accessing electronic documents is how NAL is
connected to the outside (Internet), and how we are connected to each other
within NAL. Depending on the kind of technology we have to process and access
the electronic media, there are different scenarios for how it will be done.

A.   If stand-alone workstations are in each of the Technical Services and
     Public Services Branches, there will be no simultaneous access to these
     documents.  There will be a delay as the CD-ROM winds its way through
     the Technical Services Division workflow.  DDSB will still have to chase
     down publications if there is a document delivery request for an item on
     one of the CD-ROMs. After the disk is processed, DDSB might have to
     share a single copy with Reference or the Software Demonstration Center.
     If these documents are made available on a stand alone workstation, SDC
     and/or DDSB would have to purchase a workstation that will accommodate a
     number of different retrieval software packages. Each time a CD-ROM
     needs to be accessed, if the retrieval software does not reside on the
     disk itself, it would have to be loaded onto the workstation. Certain
     retrieval software systems might be used for more than one CD-ROM title,
     and those might be made available on the system at all times (for
     example, Silver Platter or K-Ware). 
   
B.   If there is an NAL-wide network on which CD-ROMs could be loaded and
     accessed. (This is assuming that we would have a site license for
     multiple simultaneous access). As items are acquired, CD-ROMs can be
     loaded on a network that will make them simultaneously accessible for
     check-in of serial titles, cataloging and indexing.  This material will
     be also be accessible for document delivery purposes before technical
     services processing is completed and this can be done with less
     disruption to all concerned if such a network exists.  After disks are
     processed, DDSB can retrieve requested publications without the
     intervention of the SDC staff.  DDSB will not have to be concerned about
     loading disks and retrieval software and DDSB can be more concerned with
     training its own staff or its contractors on how to access documents.

C.   If these publication are loaded on the ISIS computer.  If ISIS
     workstations are equipped with "windows," bibliographic, indexing, or
     holdings records can be created and/or updated on the same workstation
     where the documents themselves are viewed.  To further automate the
     processing of materials, if SGML labels are coded on documents and if
     these tags are mapped to MARC tags, this can expedite the availability
     of these materials to patrons. Again, the possibility of DDSB being able
     to access publications even if they are in-process would be an
     improvement over the current workflow.  If documents are made available
     from ISIS, patrons will have transparent access from searching the ISIS
     database to actually accessing documents from the same workstation. ISIS
     terminals already in the DDSB work areas could be used for document
     delivery purposes. DDSB may need to install one or two more ISIS
     terminals or NeXT workstations to make this possible. It is also
     possible that patrons can do their own printing, thus taking DDSB out of
     the picture entirely for some of the requests.


Conclusion

Although the recommended policies and procedures in this document are
conservative in nature, they may not be valid in one year or five years. 
Technology is changing so rapidly that it may