COMMUNITY-BASED OPEN SOURCE

The Phenomenon and Research Opportunities

Manlu Liu and Qiang Tu

Rochester Institute of Technology, New York, U.S.A.

Keywords: Community source, System development, Collaboration, Virtual team, Kuali.

Abstract: Community-based open source (community source) development has emerged as a new way of developing

enterprise applications, leading to a unique type of open source practice involving collaboration from

multiple organizations in a virtual environment. In this paper, we introduce the concept and motivation of

community source and address several research directions in community source under multiple perspectives.

We examine a real world case, the Kuali community source project, to help better understand this new

phenomenon. The long term research objectives are discussed at the end of the paper. This paper facilitates

a general understanding of the emerging community source development landscape. We believe that the

issues presented in this paper can attract more researchers to study this new area and help organizations

make better decisions in IT investments.

1 INTRODUCTION

Due to an increasingly competitive business

environment, organizations demand customized

application software that can meet their specialized

and strategic requirements. Sometimes organizations

feel compelled to develop the software in-house

since they could not find the suitable application

software in a commercial market. According to the

study of (Perry et al. 2007), the overall cost of in-

house development is so much higher than that of

buying commercial software that many

organizations cannot afford the former approach.

Furthermore, it is impractical for many organizations

to achieve all the necessary competence for building

software themselves. As a result, they seek out

strategic alliance to jointly develop their desired

software. Strategic alliance is a formal relationship

between two or more parties to pursue a set of

agreed upon goals or to meet a critical business need

while remaining independent organizations

(Yoshino et al. 1995). A new approach of enterprise

application software (EAS) development is applying

strategic alliance to the development mechanism of

open source software (Agerfalk et al. 2008). The

approach is referred to as community-base open

source or “community source” (Wheeler, 2004). In

community source, a community of partner

organizations invests and develops the EAS and this

software is open source eventually.

The community source is a unique form of open

source approach. It is somewhere between

commercial software and open source. As described

by Raymond (2001), when developing commercial

software, vendors are organized as builders

constructing cathedrals based on blueprints; but on

the other hand, open source software is built when

programmers work in a way similar to merchants in

a marketplace, harnessing a range of approaches and

agenda, taking input from diverse types of people

scattered across the world, and being open to new

ideas and new participants. A community source

organization can be described as a shopping mall

where a virtual organization consisting of

participating institutions, each of which commands

its own employees. Community source creates a

formal virtual organization where multiple partner

institutions collaborate with each other to develop

custom software solution in different locations and

at different time.

In community source, rather than relying on

commercial software vendors, the development

partner institutions pool their resources together to

develop EAS. This pooling of resources gives the

organizations better control over the software

development and community source thus offers an

innovative approach to developing open-source

EAS.

419

Liu M. and Tu Q..

COMMUNITY-BASED OPEN SOURCE - The Phenomenon and Research Opportunities.

DOI: 10.5220/0003491104190424

In Proceedings of the 13th International Conference on Enterprise Information Systems (ICEIS-2011), pages 419-424

ISBN: 978-989-8425-56-0

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

By becoming a development partner in

community source, the institution can influence the

application features, complete the deployment

sooner, minimize the total cost of ownership, and

reduce the uncertainty of the system value. This

incentive is the main driving force for institutions to

make community source projects succeed. The

motivation for institution joining in community

source development will be further explained in the

Kuali case section.

It is worthwhile to study community source since

it offers an innovative way to develop enterprise

applications and it can potentially break the

dominance of commercial software in enterprise

applications.

This paper is structured as follows: we start with

introducing a real world community source project

to help better understand the community source

approach. Then we outline the related research

directions under the technological, the economic and

the managerial perspective respectively and discuss

the potential research opportunities. We present the

future trend of community source and the value of

research in community source in the discussion

section. The long term objectives for community

source research are pointed out at the end of this

paper.

2 THE KUALI CASE

In order to better understand community source, we

studied a real-world ongoing community source

project of called Kuali (www.kuali.org). The Kuali

case offered us a great opportunity to study the

research issues of community source in a higher

education setting.

Kuali is a consortium of universities to develop

an open source financial service system, starting

with the conversion of the Indiana University’s

Financial Information System to the web in 2004

(www.kuali.org). The original motivation of the

project is that existing financial systems used in the

universities are outdated and too difficult to

maintain. The commercial products are often too

expensive and hard to customize; some institutions

paid tens of millions to ERP vendors for software

and installation, but still need to build 15% of all the

features they need to handle specific financial

transaction needs. Some schools and colleges found

that they need to operate expensive "shadow

systems" to provide needed features that are absent

in currently available ERP packages. The

alternatives to buying off-the-shelf packaged

financial software are equally daunting and can only

be considered by the largest institutions. The Kuali

project provides an attractive alternative to the “buy

or build” dilemma. It pools institutional resources to

develop an open source financial system, thus

dramatically reducing the cost of managing fiscal

data and processes in higher education.

The initial mission of the Kuali consortium was

to develop a baseline system for financial services

and has now expanded to other systems such as

research administration and student management.

Currently, there are three main application modules

in Kuali: the Kuali Financial Services (KFS), the

Kuali Research Administration (KRA), and the

Kuali Student System (KS). The KFS project was

just completed. The KRA and KS project are

scheduled to be released after 2010.

The Indiana University and the University of

Hawaii started to lead the effort to build KFS in

2004. In March, 2005, Kuali project got a start-up

grant $2.5M from the Andrew W. Mellon

Foundation. The follow-up funding was invested

from participating institutions which are referred as

the development partners of KFS. KFS project is

mainly funded by its partner institutions. In April

2006, Kuali project announced the availability of the

Kuali Test Drive, which enables institutions to

explore KFS. In October, 2006, Kuali foundation

released KFS phase I. KFS Phase II was released in

July 2008. KFS project was completed in 2010 and

have grown to around thirty development partners

including Carnegie Mellon University, Cornel

University, Massachusetts Institute of Technology,

the University of Arizona, and University of

California. One of the very important components of

the Kuali effort is the Kuali Commercial Affiliates

program. Commercial Affiliates provide for fee

guidance, support, implementation, and integration

services related to the Kuali software. Affiliates may

offer packaged versions of Kuali that provide value

for installation or integration beyond the basic Kuali

software. Affiliates may also offer other types of

training, documentation, or hosting services.

The development team in community source is

under a formal organizational structure. Figure 1

shows the complexity of the organization structure.

The development partners work in a project

organization. The Kuali Board is the final decision

maker during the development of Kuali. The Kuali

functional council and Kuali technical council take

care of the functional issues and technical issues of

Kuali, respectively. The Extended board, Kuali

investors, Kuali functional council, Kuali technical

council and the project manager report to the Kuali

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

420

board. The project manager supervises the project

staff and development staff. The project manager is

managing an organization that is virtual in every

aspect. The task of the project manager is to

complete a project of size and complexity on time

and on budget.

Figure 1: KFS Organizational Structure.

In order to ensure flexibility and extensibility,

Kuali uses the most up-to-date approaches and

technologies such as open source, workflow, and

service-oriented architecture.

As we can see from this case study, the Kuali

project provides an attractive alternative to the “buy

or build” dilemma. The Kuali project pools

resources from multiple universities together to

develop an open source system which can meet

universities’ special requirements, and finally

reduces costs and minimizes risks. In the Appendix,

we compare community source with the other two

main approaches: traditional open source and

commercial software and summarize their

characteristics. Comparing with the traditional open

source approach, Kuali has better system quality

control since the system is developed under a formal

organization and the system adopters could have

strong technical support due to the commercial

affiliates in Kuali. Comparing with commercial

software, Kuali satisfies the special requirements for

the high education and dramatically reduce the

development costs and development time by sharing

resources among multiple institutions.

3 RESEARCH DIRECTIONS

In a community source project similar to Kuali,

multiple institutions make significant investment in

developing the system. Development teams acquire,

maintain, analyze, and formalize software

requirements from multiple stakeholders that can

have varying and possibly conflicting needs. The

research directions in community source could be

generated from multiple perspectives.

We outline several research issues for

community-based development of enterprise

applications under three perspectives: the

technological perspective, the economic perspective

and the managerial perspective. We discuss unique

research opportunities along with those research

directions.

3.1 The Technological Perspective

We address two research issues under technological

perspective, including technology flexibility,

workflow modelling paradigms.

3.1.1 Technology Flexibility

Community source is a unique form of open source

in which the community must balance the various,

sometimes even conflicting, requirements from all

development partners. As such, community source

is more challenging than a general open source

project in which variations in requirements are not a

major issue (Liu, Wang and Zhao, 2007). In our

view, community source has more stringent

requirement in flexibility in order to deal with

complex requirement analysis and change

management. Recently emerged technologies such

as web services, service-oriented architectures, and

workflow automation can help make a system more

customizable. For example, Kuali takes advantage

remarkably of these emerging technologies to enable

the strong system flexibility. In fact, we believe that

community source is a good representative of a

recent trend in combining service centricity with

open source development to build scalable and

flexible information systems. As such, how

technology flexibility is achieved through service-

centric community source becomes a very

interesting research issue.

3.1.2 Workflow Modelling Paradigms

Currently, the Kuali platform specifies workflows by

means of workflow rules in the context of financial

documents. As such, these workflow rules are also

called routing rules. These routing rules seem to

work fine in the context of financial services since

the workflows are relatively simple with typically

three to five steps. However, it is not clear if routing

rules would be appropriate in the modules which

consists of many processes (Liu, Wang and Zhao,

2007). Specifying these workflows using routing

rules could result in lengthy XML scripts that are

COMMUNITY-BASED OPEN SOURCE - The Phenomenon and Research Opportunities

421

very difficult to write and maintain. In addition, the

control flow of those processes is not explicitly

modeled when using rule-based and document-

driven workflow modeling, which makes it difficult

to understand the underlying process logic. This

raises a research issue on when a particular

workflow modeling paradigm should be used.

3.2 The Economic Perspective

The community source approach provides a viable

alternative to vender-provided packaged solutions by

combining effectively the benefits of in-house

development and outsourcing. One critical decision

for prospective community source partners is

whether or not it should pay a significant amount of

partnership fee to join the development community,

which we refer to as “the community source

investment decision” or “the community partnership

decision”(Liu, Zeng and Zhao, 2008). As in general,

the resulting software package of a community

source project is available as an open source free of

charge, it seems counter-intuitive for an organization

to be willing to make major investments. As shown

in many real-world examples, many organizations

have already invested heavily in various kinds of

community source projects. This poses an interesting

and timely research challenge as to gaining an in-

depth understanding of economic incentives behind

these decisions and further developing an actionable

decision framework for the community partnership

decisions.

3.3 The Managerial Perspective

In community source, multiple institutions

collaborate closely and work in a virtual

environment to develop the system. Eventually the

system needs to meet the requirement from multiple

institutions. We can see that project management

issue becomes a critical issue of developing the

system successfully in community source.

We address two research issues under managerial

perspective: project success, outsourcing.

3.3.1 Project Success

In community source, multiple institutions make

investments and develop systems together. The

whole development process requires close

collaboration among partner institutions. The

developers in community source are full time

employees in partner institutions, who are

designated to work for the community. The

community source organization is a virtual

organization consisting employees from the partner

institutions. Community source development is

therefore very different from in-house software

development and traditional open source

development. Achieving project success in

community source poses new challenges that are not

found in the development of commercial software or

conventional open source software such as Linux

(Liu, Wheeler and Zhao, 2008).

Since community source is a new approach of

system development, understanding how to achieve

project success in community source development is

a very interesting research topic. However, the

related issues have not been well studied in the

literature. Several key research questions motivate

our work: (1) which framework can be used as a

theoretical basis for studying project success in

community source development? (2) What factors

can be identified within the theoretical framework?

(3) How do these factors affect project success in

community source development?

3.3.2 Outsourcing

In community source, rather than rely on

commercial software vendors, the user organizations

pool their resources together to develop EAS. This

pooling of resources gives the organizations better

control over the software development and

community source thus offers an innovative

approach to developing open-source EAS.

Developers in a community source project are

employees of the partner institutions who are

designated to work on the project. Due to this unique

feature, however, community source projects face a

number of challenges. On one hand, it is optimal for

the community source project to attract as many

participating organizations possible in order to share

resources, reduce costs, and minimize risks. On the

other hand, the management of community source

project becomes increasingly complicated and

difficult with the increasing number of participating

institutions.

One of the solutions to those issues might be

global outsourcing of software development in the

community source project (Liu et al., 2010). The

primary research questions of this study could be:

What are the motivations for global outsourcing of

software development in community source, and

what are the potential benefits of global outsourcing

of software development in community source?

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

422

4 DISCUSSIONS

The greatest value of community source is the

leveraging of resources of the partners and the

community for shared value creation. In the past,

when Cornell University spent $500,000 for some

system, the investment provided no advantage to

San Joaquin Delta College; Or when Rutgers

University developed a clever piece of cross-

language middleware code, Indiana University did

not benefit; When Cambridge University developed

a teaching tool, U.S. and other institutions gained

little (Wheeler, 2007). The community source

approach changes the situation in which

organizations invest huge money in IT but do not or

seldom share among each other. All of organizations

in community source are mutually using other

people’s money to get and sustain the systems they

need. Community source are foremost a

coordination mechanism for institutional, corporate,

and personal investments of resources, ideas, and

talent.

Although community source faces lots of

uncertainty and might not become popular in the

long term for some reason. For example, the

commercial software might decide to lower the price

due to the competition from the community source.

However, as long as community source makes the

impact to the society, the research in community

source has the unique value to the application

software development and the result of studies will

make contribute to both theory and practice. The

collaborative approach to solving problems faced by

institutions with an open source license has the

highest probability of benefiting the most peoples

and organizations. The essence of community source

approach is that the institutions develop the

collaborative capability that extracts the greatest

value from open source.

5 LONG TERM OBJECTIVES

Since community source becomes a new approach

for enterprise application development, there is a

clear need for a theoretically rigorous and practically

relevant framework for understanding value create

and recreate through enterprise application

development under community source environment.

The long term research objectives should be

proposing the enterprise application development

cycle as a collaborative capabilities theory for

measuring, predicting, and understanding an

institution’s ability to create value for community

through enterprise application development.

Eventually, a generic framework should be

developed for describing the process of acquiring

application software via either commercial, open

source, or home grown approach.

REFERENCES

Agerfalk, P. J., Fitzgerald, B., 2008. Outsourcing to an

Unknown Workforce: Exploring Opensourcing as a

Global Sourcing Strategy. MIS Quarterly (32:2), June,

pp 385-409.

Liu, M., Zeng, D. D., Zhao, J. L., 2008. A Cooperative

Analysis Framework for Investment Decisions in

Community Source Partnerships. Proceedings of the

Americas Conference on Information Systems, August

14 - 17, Toronto, Canada.

Liu, M., Wang, H. J., Zhao, J. L., 2007. Achieving

Flexibility via Service-Centric Community Source: the

Case of Kuali. Proceedings of the Americas

Conference on Information Systems, August 9-12.

Liu, M., Wheeler, B., Zhao, J. L., 2008. On Assessment of

Project Success in Community Source Development.

Proceeding of International Conference on

Information Systems.

Liu, M., Wu, X., Zhao, J. L., Zhu, L., 2010. Outsourcing

of Community Source: Identifying Motivations and

Benefits. Journal of Global Information Management,

18(4).

Perry, R., Quirk K., 2007. An Evaluation of Build Versus

Buy for Portal Solutions. IDC White paper, sponsered

by IBM, December.

Raymond, E., 2001. The Cathedral and the Bazaar.

Sebastopol, CA: O.Reilly Media., Inc..

Wheeler, B., 2004. The Open Source Parade. EDUCAUSE

Review, pp. 68-69.

Wheeler, B., 2007. Open Source 2010: Reflections on

2007. EDUCAUSE Review, pp. 49-52.

Yoshino, M. Y., Rangan, U. S., 1995. Strategic Alliances

– An Entrepreneurial Approach to Globalization. First

Edition. Harvard Business School Press.

COMMUNITY-BASED OPEN SOURCE - The Phenomenon and Research Opportunities

423

APPENDIX

Comparison among Community Source, Traditional Open Source and Commercial Software

Community Source Traditional Open Source Commercial Software

Examples

Kuali MySQL SAP

Builders

Organized teams of formal

employees from multiple partner

institutions

Loosely connected volunteer

developers

System development departments

in software companies

Users

Development partners, deployment

members, and others

Organizations or individuals who

use the software for free

Organizations or individuals who

pay for the software

Characteristics

The application development requires

initial investments which are shared

by the partners; the result is open

source.

The software is free for adopters;

the developers are volunteers.

The users need to buy the

software; dedicated investment is

required to develop the software.

Initial

investment

Cost sharing among development

partners

Low Purchasing price

Customization

The system is designed to meet the

requirements of development partners

The system is designed to meet

some generic requirements

Standard package

Quality control

Strong Weak Strong

Development

and deployment

cycle time

Short: system is design to meet user

requirements in the first place;

developers from partner institutions

are well trained from the

development process.

Long: due to uncertainties in

development process and required

system customization after

adoption.

Long: due to required system

customizations after adoption.

Technical

support after

adoption

Strong Weak Strong

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

424