United States
General Accounting Office
Washington, D.C. 20548
Information Management and Technology Division
B-251542
January 28, 1993
The Honorable John P. Murtha
Chairman, Subcommittee on Defense
Committee on Appropriations
House of Representatives
Dear Mr. Chairman:
The Department of Defense (DOD) estimates that expenditures for
developing and maintaining software for its weapons, command and
control, and other automated information systems currently exceed $24
billion a year. In an attempt to better manage these costs and improve
its ability to develop and maintain high quality software, Defense has
initiated a comprehensive effort to incorporate software reuse practices
into its software development efforts. Software reuseÑthe practice of
developing new applications from existing softwareÑoffers the potential
to greatly reduce the time, cost, and effort needed to develop and
maintain high quality software.
As requested by your office, this report provides background information
on software reuse, including an overview of issues that can inhibit
effective software reuse and information on Defense's strategy to
implement a departmentwide software reuse program. Appendix I further
details our objectives, scope, and methodology. Appendix II provides
information on Defense's initiatives to incorporate software reuse into
its software development process.
Results in Brief
Developing and maintaining software in organizations such as the
Department of Defense is very costly. According to many experts in the
software community, software reuse is a possible solution to reduce
these costs, as well as to increase software productivity and
reliability. Although these benefits and savings are compelling,
achieving, them will require the resolution of significant technical,
organizational, and legal issues.
Even while proclaiming the potential of reuse, many software experts
have questioned the maturity of software reuse. These experts indicate
that methodologies to implement reuse have not been fully developed,
tools to support a reuse process are lacking, and standards to guide
critical software reuse activities have not been established.
Page 1 GAO/lMTEC-93-16 Issues Facing Software Reuse
B-251542
Beyond such technical difficulties, organizations also face numerous
challenges to effectively implement and practice software reuse. An
organization must make a significant commitment to reuse because
fundamental changes in the organization's software development approach
will be needed and significant up-front costs for training and tools
will be required. Further, uncertainties in legal policies, such as
liability and intellectual property rights that currently hinder
software reuse, need to be addressed, and acquisition policies need to
be modified to better promote reuse.
Background Software reuse is the practice of using existing software
components to develop new applications. Reusable software components can
be executable programs, code segments, documentation, requirements,
design and architectures, test data and test plans, or software tools.
They may also be knowledge and information needed to understand,
develop, use, or maintain the component. Figure 1 shows examples of the
different types of reusable software components.
Page 2 GAO/lMTEC-93-16 Issues Facing Software Reuse
B-251542
Figure 1: Examples of Reusable Software Components
Code
|
Executable | Documentation
programs \ | /
\ | /
\ | /
Software ------------ Reusable --------- Software
tools Software requirements
/ Components
/ | \
Knowledge and / | \ Designs and
experience | \ architectures
|
Test data and
test plans
There are two basic forms of software reuseÑopportunistic and
systematic. Opportunistic reuse is practiced in an ad-hoc fashion during
software development. In opportunistic reuse, new applications are
developed from software that has been salvaged from existing systems and
modified to meet the specific needs of that application. Systematic
reuse is planned and integrated into a well-defined software development
process. In systematic reuse, new applications are developed from
software that has been designed and developed to be reused specifically
for other similar applications.(1)
(1) In systematic reuse, new reusable software is also created as a by-
product of applications development.
Page 3 GAO/lMTEC-93-16 Issues Facing Software Reuse
B-251542
Software reuse can be practiced vertically or horizontally. Vertical
reuse is the reuse of software components within a single domain.(2) For
example, a software component that implements procedures to withdraw
federal taxes from a paycheck can be reused by different accounting
systems within the payroll application domain. Horizontal reuse, on the
other hand, is the reuse of software components across different
domains. For example, software components, such as sort and merge
procedures, can be reused by systems in many application domains.
Software Reuse Process
The software reuse process consists of three stages: component creation,
component management, and component utilization, as shown in figure 2.
Figure 2: Conceptual Framework of the Software Reuse Process
Software Reuse Process
-----------
|Component|
|Creation |--->-- Components ---
---------- |
^ |
| v
| --------------
Feedback ---<----<- | Component | --------> Components
| | management | |
| -------------- |
^ |
| ---------------
| | Component |
|----- Feedback -----<-----------<----- | Utilization |
---------------
|
v
Application software
systems and other
software-related
products
Source: DARPA
(2) A domain is a family of related systems that exhibit common objects
and operations.
Page 4 GAO/lMTEC-93-16 Issues Facing Software Reuse
B-251542
This framework, established by the Defense Advanced Research Projects
Agency's (DARPA) Software Technology for Adaptable Reliable Systems
(STARS) program, presents the flow of information within the software
reuse process and its products. During component creation, domains where
reuse is possible are identified and reusable software components are
developed. Once components are developed, they are stored and managed in
a software repository, which is a library that allows users to access,
search, and retrieve the components. Key functions of component
management include certifying, classifying, and cataloguing components,
as well as configuration control of the software components as a result
of software upgrades and maintenance. Figure 3 illustrates the basic
functions of component management.
Figure 3: Key Functions of Component Management
Reusable Certify component for Classify component
software ----> quality, functionality ----> according to its
component and completeness characteristics
|
|
|
----------------- V
| | Catalogue component
| Repository | <---- and any related
| | information
-----------------
^ |
| |
| V
Maintain component
and control its
configuration
In component utilization, components in the repository are searched,
evaluated, selected, and integrated into the software product under
development. Components can be used to either develop application
software systems or create new reusable components and software-related
products.
Potential Benefits of Software Reuse
Systematic reuse is viewed as a possible means to reduce software
development costs while improving software quality. According to a
number of software experts, reuse has the potential to
Page 5 GAO/lMTEC-93-16 Issues Facing Software Reuse
B-251542
- increase productivity by reducing the time and effort needed to
develop software,
- increase reliability because systems will be developed with thoroughly
tested and proven components,
- reduce costs by sharing knowledge and practices needed to develop and
maintain software, and
- establish a more standard and consistent approach to software
development and maintenance by using common components and procedures.
As an example, the Software Engineering Laboratory (SEL) at the National
Aeronautics and Space Administration's (NASA) Goddard Space Flight
Center achieved significant benefits by implementing software reuse in
the development of software products in its Flight Dynamics Division. In
a 1991 study, SEL reported
- a 35-percent reduction in effort needed to deliver a line of code
(from .65 to .42 staff hours),
- a 53-percent increase in daily productivity (from 12.4 to 19 lines of
code per day), and
- an 87-percent increase in quality (from 3.9 errors to .5 errors per
thousand lines of delivered code).(3)
While the results of the SEL study appear promising, experts caution
that the benefits of software reuse are not easily or quickly realized.
The potential impact of software reuse remains questionable because of
technical, organizational, and legal issues that need to be addressed.
Technical Issues Establishing a systematic software reuse program is
difficult. Few organizationsÑin either the private or public sectorsÑ
have been able to incorporate software reuse into their software
development practices because the technical knowledge to develop and
apply software reuse methodologies, standards, and tools is still
evolving. Table 1 summarizes the technical barriers to software reuse
discussed in the following sections.
(3) Proceedings of the Sixteenth Annual NASA/Goddard Software
Engineering Workshop: Experiments in Software Engineering Technology,
Software Engineering Laboratory, December 1991.
Page 6 GAO/lMTEC-93-16 Issues Facing Software Reuse
B-251542
Table 1: Summary of Technical Issues
Domain analysis
-lack of standard methods to process information on
domains
-lack of standard methods to represent the outputs of
domain analysis
Classification of software components
-no accepted standards to classify
components
-classification depends upon a domain analysis
Interoperability of software repositories
-lack of standards for interoperation of repositories
Adaptation of software components
-adaptation depends upon the availability of
information on component
-more required adaptation can offset the savings and
benefits of software reuse
Reuse of systems designs and architectures
-designs and architectures are harder to represent
because they are more abstract
-lack of standards to represent designs and
architectures
-lack of tools to represent, develop, and maintain
designs and architectures
Software metrics -lack of standard metrics
-inconsistent interpretation of metrics
-collecting metrics is expensive and time-consuming
Domain Analysis
Domain analysis involves systematically gathering and representing
information on software applications. Experts in the software community
generally agree that domain analysis is at the "heart of reuse." Its
purpose is to generalize common features in similar application areas,
identify the common objects and operations m these areas, and define and
describe their relationships. Once collected, the information can be
used to create reusable software components that support these areas.
For example, in an airline reservation system domain, common objects are
flights and seats, while common operations include flight scheduling and
seat assignments. These objects and operations are related in specific
ways to the airline reservation system domain. As such, software
components that support these objects and operations could be reused by
developers of other airline reservation systems.
Domain analysis is a complex process that involves acquiring and
representing knowledge on specific domains. Information on the domain
must be identified, compiled, analyzed, and represented in a format so
that it can be reused. The domain analyst needs to not only identify the
objects and operations and their relationships in the domain, but also
be able to explicitly represent that information so others can easily
understand and reuse it.
Page 7 GAO/lMTEC-93-16 Issues Facing Software Reuse
B-2C1642
However, standard methods to process and represent information on a
domain are lacking. Current domain analysis methodologies, such as the
Software Engineering Institute's Feature Oriented Domain Analysis (FODA)
and Dr. Ruben Prieto-Diaz's Top-Down, Bottom-up Approach, are still
evolving and thus do not completely address these functions.(4)
Classification of Software Components
Classification is a process of systematically grouping reusable software
components stored in a software repository. A classification scheme for
a software repository is analogous to the Dewey Decimal System for a
library. Its purpose is to provide the basic organization of a
repository so users can easily access, search, and retrieve components
in the repository.
Establishing a classification scheme is knowledge-intensive and time-
consuming. It requires combining the knowledge inherent in the
components of the repository with the knowledge about the application
domain where the components are going to be used. Common characteristics
of the components are then grouped and organized into a structure that
can be easily understood by repository users. While automated tools
exist to catalogue software components (store and retrieve components in
a repository), the key difficulty in classification is how to organize
the overall repository because there are no accepted standards for
classifying components.
Interoperability of Software Repositories
Interoperability is the ability of two or more systems to exchange
information- It is an important capability in instances where multiple
repositories exist because it permits software repositories to share
components, reduce the number of redundant components in the different
repositories, and make components available to all repository users.
Development is currently underway, for example, in DARPA's STARS program
to establish an architectural framework for repository interoperability.
However, standards for interoperability of software repositories, such
as nomenclature, communication protocols, and component exchange
formats, do not exist. Currently the Reuse Library Interoperability
Group (RIG) is addressing standards for interoperability and plans to
submit
(4) For more information on FODA, contact Dr. Sholom Cohen, Software
Engineering Institute, Carnegie Mellon University, Pittsburgh, Pa For
more information on the Diaz model, contact Dr. Ruben Prieto-Diaz, Reuse
Inc., Fairfax, VA
Page 8 GAO/lMTEC-93-16 Issues Facing Software Reuse
B-251542
proposals to standards organizations, such as the Institute of
Electrical and Electronics Engineering.(5)
Adaptation of Software Components
Adaptation involves modifying a software component to make it reusable
in different software applications. It requires the software developer
to determine what interfaces are needed and then tailor the component
and/or application to make them operate together. Since the current
state-of-practice is mainly opportunistic, most of the benefits that can
be gained from software reuse are highly dependent on effective
adaptation methods. However, adaptation is a difficult process because
the developer has to understand
- how the component currently functions,
- how the new application works, and
- what modifications are needed to make the component work in the new
application.
Without this information, a developer cannot easily adapt the software
component for reuse. Even with the information, the adaptation process
can be labor-intensive, potentially offsetting time and cost savings
promised from software reuse.
Reuse of Systems Designs and Architectures
Although the current state-of-practice of software reuse has been mainly
limited to the reuse of code, experts believe that the reuse of other
software products, such as systems designs and architectures, can
further increase the benefits of software reuse. They call this "higher-
level reuse" because it involves reusing products that are from software
development phases that occur prior to (or higher than) the one in which
code is written.(6) According to these experts, the reuse of higher-
level components will yield greater benefits because designs and
architectures
- are more flexible than code because they are independent of language,
hardware platforms, and implementation-specific details
- represent application solutions rather than implementation solutions;
and
(5) RIG is a volunteer organization composed of members from government,
academia, and private industry. Membership is open to any organization
interested in the interoperability of government-sponsored reuse
libraries.
(6) In the traditional software development process, there are four
successive phases: planning, design, coding and testing, and integration
and testing. In the planning phase, requirements are set; during the
design phase, designs and architectures are developed; in the coding and
testing phase, code is written and tested; and in the integration and
testing phase, the coded components are combined and tested as a whole.
Page 9 GAO/IMTEC-93-16 Issues Facing Software Reuse
B-251542
- can be used to automatically create lower-level components, such as
code.
However, formally representing systems designs and architectures in a
reusable form is very difficult because they are not as tangible as
code. Further, standards and tools to represent and develop systems
designs and architectures are lacking.
Software Metrics
Software metrics are quantifiable measures that are used to assess the
products and processes of software development. Such metrics may include
measures of usefulness, cost, and quality that could be used to better
manage software development programs. However, identifying and
establishing metrics is difficult because standard methodologies do not
exist to collect data for software development and products in general,
or for reuse in particular. As such,
- current metrics are inconsistent,
- interpretation of metrics can vary from individual to individual, and
- collecting metrics is a very expensive and time-consuming process.
Without effective metrics, organizations cannot adequately determine the
costs and benefits of incorporating software reuse into their software
development processes.
Organizational Issues
Software reuse will not happen merely because the technical means for
achieving it become available. Software experts told us that top
management must be convinced to make a business decision to incorporate
systematic reuse into the software development process. Further, project
managers and software developers must be willing to make fundamental
changes in the way they develop software. Otherwise, software reuse will
remain at the opportunistic level, and the potentially greater benefits
of systematic reuse will not be realized.
Gaining Management Support and Commitment
Experts have stated that for a software reuse program to be successful,
top management must decide and commit to implementing a systematic reuse
program throughout the organization. They noted that top management
needs to
- incorporate software reuse practices into the software development
process,
Page 10 GAO/lMTEC-93-16 Issues Facing Software Reuse
B-251542
- train and educate employees on software reuse,
- develop and provide tools to practice software reuse, and
- allocate the proper funds and resources to support a reuse program.
However, experts generally agree that overall such management support is
lacking. For example, at a jointly sponsored workshop by the Software
Productivity Consortium, the Microelectronics and Computer Technology
Corporation, the Software Engineering Institute, and the Rocky Mountain
Institute on Software Engineering, attendees unanimously agreed that
management generally has a shortsighted view on software development and
is often not willing to commit resources to acquire needed tools and
training in software reuse technology.
In addition, some experts believe that top management is hesitant to
invest in software reuse because the benefits of software reuse are not
quickly realized and are uncertain. To illustrate, some experts estimate
that the savings from reusing a component will not be realized until
that component has been reused at least three times and believe that it
initially costs about 20 to 55 percent more to develop reusable
software.
Gaining Support of Project Another common organizational issue is the
unwillingness of project
Managers and Developers managers and software developers to develop and
use reusable
components. As noted above, developing reusable software is more costly
and time-consuming. As such, project managers, who are often pushed by
limited funds and tight schedules, have little incentive to allocate the
additional time and resources needed to develop reusable software
components.
Additionally, software developers are often reluctant to accept and use
reusable components for fear that the components will not be as
efficient, effective, or reliable as the software they write. Further,
using reusable software components requires that the components be
understood and adapted to meet the specific needs of a software system
before it can be integrated. In either case, the reluctance of software
developers to use reusable software and the lack of incentives for
program managers to develop reusable software components remain issues
that need to be addressed.
Legal Issues
The software community s hope for widespread reuse also brings about a
number of challenging legal issues. The Institute for Defense Analysis
(IDA)
Page 11 GAO/lMTEC-93-16 Issues Facing Software Reuse
B-261642
recently published the proceedings of a 1990 workshop on legal issues,
sponsored by the Strategic Defense Initiative Organization.(7) The
workshop reached the following conclusions:
- Large-scale software reuse will likely cause more complex and more
frequent encounters with legal issues.
- Large-scale reuse will require a national registry that tracks the
source of original development and modifications for each component.
- A mechanism is needed to reward developers who modify and add value to
existing components, while still protecting the rights of the original
developer.
- It is not good policy and may conflict with federal law if the
government assumes all legal liabilities associated with a reuse
repository.
- Software patents and licensing policies need to be addressed.
Our discussions with reuse experts in the software community
corroborated these concerns. Most believed that strategies are needed to
address intellectual property rights, liability, and acquisition
policies of reuse.
Intellectual Property Rights
Software is protected legally as intellectual property through laws that
control its dissemination and use. These laws relate to the exclusive
ownership of the idea, the form of expression of the idea, and the use
of the idea and its expression. There are three basic methods to protect
software: patents, copyrights, and trade secrets. Patents protect the
rights to the idea itself, while copyrights protect the rights to the
expression of the idea. Trade secret laws protect the rights to
confidential business information.(8) However, in many cases laws are
not clear about the enforcement of intellectual property rights. As
such, a major challenge facing software reuse is to balance these rights
between software suppliers, repositories, and users.
Several approaches have been proposed by various members of the software
community to address intellectual property rights. For example, one
approach proposed having repositories acquire full rights to software
components. However, questions were raised about a repository's ability
(7) Proceedings of the Workshop on Legal Issues in Software Reuse (IDA
Document D-1004), Institute for Defense Analysis, July 1991.
(8) Patents and copyrights are governed by federal law. Trade secrets
are governed by state laws that may vary from state to state.
Page 12 GAO/IMTEC-93-16 Issues Facing Software Reuse
B-251542
to motivate suppliers to do this because the component supplier will
lose exclusive rights to commercially market the component.
Another approach proposed having software suppliers license limited
software rights to a repository. However, issues were raised that if
rights are transferred through a licensing agreement, future users of
the component would need to be protected from breaching license
agreements. Further, repositories would need to track all uses of
software components to ensure that royalties and service fees are
compensated to the component supplier and that licensing agreements are
enforced. If the repository is unable to track software components and
enforce licensing agreements, suppliers could be discouraged from giving
up partial software rights to the repository.
Liability
Liability, in the context of software, refers to the legal
responsibility for harm attributable to software components. Liability
may affect not only the supplier of the component, but also the
repository and user. For example, software suppliers could be liable for
submitting defective software components that fail to meet performance
standards or cause a software system malfunction. Repositories could be
liable for marketing and distributing defective components or not
properly enforcing the rights to software components. Users could be
liable for infringing the intellectual property rights attached to a
software component.
However, the subject of liability for software is fairly new to the law.
As a consequence, there are still questions, such as whether software is
a product or service, that have left some uncertainty about the nature
of liability that may accompany software. For this reason, experts
believe that organizations interested in reuse need to address liability
issues, otherwise
- suppliers may be reluctant to submit components for reuse,
- repositories may limit the availability of components, and
- users may be unwilling to use the components in the repository.
Acquisition Policies
Many in the software reuse community acknowledge that changes need to be
made in federal acquisition policies of software systems before software
reuse can be effective. There are concerns that if industry is not
involved in these efforts, reuse goals will not be achieved. Some of the
issues that have arisen include how reuse should be considered in the
Page 13 GAO/lMTEC-93-16 Issues Facing Software Reuse
B-251542
request for proposals process, what criteria to use to evaluate
proposals, how costs of reuse will be evaluated and estimated, and what
incentives are needed in solicitation documents to promote reuse. These
concerns have prompted the actions of groups such as the Special
Interest Group Ada and Institute for Defense Analysis, which sponsored a
workshop to determine how and what to incorporate into the procurement
process to encourage and promote reuse.
As requested, we did not provide a draft of this report to the
Department of Defense. Instead, we discussed the facts of this report
with officials from the Office of the Director for Defense Information;
the Defense Information Systems Agency's Center for Information
Management; the Air Force, Army, and Navy; and with software experts in
industry. These officials generally agreed with the facts as presented.
We have incorporated their views in the report as appropriate.
We conducted our review between April 1992 and December 1992, in
accordance with generally accepted government auditing standards. As
agreed with your office, unless you publicly announce the contents of
this report earlier, we plan no further distribution until 30 days from
the date of this letter. We will then send copies of this report to
other interested committees; the Director for Defense Information; the
Director for Defense Research and Engineering; and other interested
parties. Copies will also be made available to others upon request.
If you or your staff have any questions concerning this report, please
contact me at (202) 512-6240. Other major contributors are listed in
appendix III.
Sincerely yours,
Samuel W. Bowlin
Director, Defense and Security
Information Systems
Page 14 GAO/lMTEC-93-16 Issues Facing Software Reuse
Contents
Letter
Appendix I 18
Objectives, Scope, and Methodology
Appendix II 19
Department of Defense Reuse Initiative
Appendix III 21
Major Contributors to This Report
Table
Table 1: Summary of Technical Issues 7
Figures
Figure 1: Examples of Reusable Software Components 3
Figure 2: Conceptual Framework of the Software Reuse Process 4
Figure 3: Key Functions of Component Management 5
Abbreviations
ASDC3I Assistant Secretary of Defense for Command, Control,
Communications and Intelligence
CARDS Central Archive for Reusable Defense Software
DARPA Defense Advanced Research Projects Agency
DDRE Director for Defense Research and Engineering
DISA Defense Information Systems Agency
DLA Defense Logistics Agency
DOD Department of Defense
GAO General Accounting Office
FODA Feature Oriented Domain Analysis
IDA Institute for Defense Analysis
IMTEC Information Management and Technology Division
NASA National Aeronautics and Space Administration
RIG Reuse Library Interoperability Group
SEL Software Engineering Laboratory
STARS Software Technology for Adaptable Reliable Systems
Page 16 GAO/lMTEC-93-16 Issues Facing Software Reuse
Appendix I
Objectives, Scope, and Methodology
On February 7, 1992, the Chairman of the Subcommittee on Defense, House
Appropriations Committee, requested that we provide background
information on software reuse, including an overview of issues that can
inhibit effective software reuse, and information on Defense's strategy
to implement a departmentwide software reuse program.
To meet our objectives, we
- met with reuse experts in private industry, government, and academia
to discuss the concepts of reuse, including benefits and issues of
effective reuse;
- attended the 5th Annual Software Reuse Workshop in Palo Alto, Ca., and
reviewed papers to identify the state of reuse;
- observed and discussed reuse experiences with private industry,
private organizations, and special interest groups;
- met with Defense officials to identify roles, responsibilities, and
strategies for Defense's departmentwide software reuse initiative; and
- examined reuse activities and research and development efforts
underway in Defense.
We performed our work at the Office of the Director for Defense
Information, Arlington, Va.; Center for Information Management,
Arlington, Va.; Defense Advanced Research and Projects Agency,
Arlington, Va.; Software Engineering Institute, Pittsburgh, Pa.; U.S.
Army Software Reuse Center, Washington, D.C.; U.S. Navy Information
Systems Management Center, Washington, D.C.; U.S. Air Force Systems and
Software Design, Hanscom Air Force Base, Ma.; Defense Logistics Agency's
Systems Automation Center, Columbus, Oh.; and National Aeronautics and
Space Administration's Goddard Space Flight Center, Greenbelt, Md.
We also visited the offices of International Business Machines
Corporation in Gaithersburg, Md., Manassas, Va., Rockville, Md., and
Owego, Ny; University of Maryland, College Park, Md.; Software
Productivity Consortium, Herndon, Va.; Reuse, Inc. Fairfax, Va.;
Raytheon Missile Systems Division, Bedford, Ma.; Mitsubishi Electric
Research Laboratories, Cambridge, Ma.; Westinghouse Electric
Corporation, Baltimore, Md.; Massachusetts Institute of Technology,
Boston, Ma; and Hewlett-Packard, Palo Alto, Ca.
Page 18 GAO/lMTEC-93-16 Issues Facing Software Reuse
Appendix II
Department of Defense Reuse Initiative
Over the last few years, software reuse has gained increased attention
throughout the Department of Defense as a way to reduce software costs
and improve productivity and software quality. A draft of Defense's
software technology strategy states that the greatest estimated Defense
cost savings over the next 16 years will come from reusing software
assets a savings of $11.3 billion in constant 1992 dollars by the year
2008. Other Defense documents note that the benefits of reuse go beyond
cost savings to include substantial increases in productivity from
avoidance of rework, and added software quality through the use of
tested components.
Responsibility for software within the Department of Defense is divided
between the Director for Defense Research and Engineering (DDR&E), who
is responsible for embedded systems and information technology research,
and the Assistant Secretary of Defense for Command, Control,
Communications and Intelligence (ASDC3I), who carries responsibility for
information systems and command and control systems.
The Defense Software Reuse Initiative
A Memorandum of Agreement between ASDC3I and DDR&E, effective November
26, 1991, established a cooperative partnership for implementing
software and other information technology initiatives for the Department
of Defense. On the basis of this agreement, the Director for Defense
Information proposed a Defense software reuse initiative to provide a
"single, consistent departmentwide software reuse strategy, with
associated policies, practices, approaches, and programs." The
initiative sought to build partnerships among users of reusable
components, suppliers of these components, and the research and
development community.
The Defense software reuse initiative is a voluntary and cooperative
alliance of individual DOD reuse activities with active participation
from the three major software reuse programs: Air Force's Central
Archive for Reusable Defense Software (CARDS), DARPA's Software
Technology for Adaptable Reliable Systems (STARS), and the Defense
Information Systems Agency's (DISA) Software Reuse Program. It is guided
by a software reuse executive steering committee representing the
ASDC3I, DDR&E, Joint Staff, Army, Navy, Air Force, DISA, Defense
Logistics Agency (DLA), Defense Intelligence Agency, and the National
Security Agency. The steering committee reports to both ASDC3I and
DDR&E, and is supported by working groups responsible for addressing
technical and management issues. DISA's Center for Information
(1) Draft of Department of Defense Software Technology Strategy,
Director of Defense Research and Engineering, December 1991.
Page l9 GAO/lMTEC-93-16 Issues Facing Software Reuse
Appendix II
Management is managing the initiative and providing a focal point for
coordination.
Initiative's Vision and Defense's software reuse initiative holds a
vision, "to drive the DOD Strategy software community from its current
'reinvent the software cycle' to a process-driven, domain-specific,
architecture-centric, library-based way of constructing software." The
strategy for achieving this vision lies in systemizing the reuse process
by identifying opportunities for reuse and establishing a process to
capitalize on those opportunities. Defense details 10 elements of this
strategy:
- Specify the domains where reuse opportunities exist and identify
criteria to prioritize, qualify, and select domains for application of
reuse techniques.
- Define the types of products suitable for reuse and develop criteria
to validate these components for new applications.
- Determine what ownership criteria pertain to these components and
require conscious decisions regarding their ownership.
- Modify the current acquisition process so reuse is integrated into
each phase of the acquisition process and into the overall
system/software life cycle.
- Define models that may suggest novel strategies and require tailored
acquisition approaches to support reuse, in order to guide business
decisions.
Establish procedures to collect metrics that (l) measure the payoff
from the reuse initiative and (2) aid developers in the selection of
reusable components.
- Define standards for the various types of components that will permit
their certification for reuse.
- Pursue a technology-based investment strategy that identifies, tracks,
and transitions appropriate reuse-oriented process and product
technologies.
- Conduct comprehensive training to ensure that practitioners and
policymakers capitalize on the initiative.
- Exploit near-term products and services that facilitate movement to a
reuse-based paradigm.
Page 20 GAO/lMTEC-93-16 Issues Facing Software Reuse
Appendix III
Major Contributors to This Report
Information Management and Technology Division, Washington DC
Frank W. Deffer, Assistant Director
David Chao, Technical Adviser
Joseph J.H. Cho, Evaluator-in-Charge
Colleen M. Phillips, Senior Evaluator
Wiley E. Poindexter,Jr., Senior Evaluator