Free and Open Source Software communities remain elusive and intangible
despite the significant amount of research that has been done on the subject. The
significance of these communities is also something that has been under much debate.
Some authors (Raymond^ 2000; Lanzara & Momer, 2003; Oh & Jeon, 2004) describe
FOSS communities as entirely virtual systems that operate almost exclusively over
the Internet on a global scale. Other authors (Krishnamurthy, 2002; O’Mahony &
Ferraro, 2004) maintain that in many cases, a significant amount of FOSS
communities often operates off-line in the ‘real world’, and that a considerable
quantity of FOSS development is actually performed by individuals. It is probable
that in actual fact, FOSS development is a mixture of both these theories. While some
projects will have large numbers of people working on them, other projects may have
few or a single developer. Furthermore, although some projects will exist entirely on-
Please use the following format when citing this chapter:
Schofield, A., and Cooper, G.S., 2006, in IFIP International Federation for Information
Processing, Volume 203, Open Source Systems, eds. Damiani, E., Fitzgerald, B., Scacchi,
W., Scotto, M., Succi, G., (Boston: Springer), pp. 221-231
222 Andrew Schofield and Professor Grahame S. Cooper
line, others may involve off-line meetings between people, especially between the
core development team and in projects originating from within organisations
(Schofield & Mitra 2005).
The community members themselves are not easily put into categories. The work
by Zhang & Storck (2001) illustrates this issue by putting forward the definition of
“peripheral members”. These are members of the FOSS community that may not
directly participate within the community. To take this idea further, the only visible
members of a FOSS community are those who participate in discussion forums,
bulletin boards, named code development, or those who make themselves known in
other ways. Members who visit the on-line communities, perhaps reading from
forums, but not posting anything, may still be considered to be part of the community
but will remain unknown to other members. In contrast to this, it is the belief of many
authors (Sagers, 2004; O’Mahony, 2004) that social interaction is the foundation to
FOSS community existence, which suggests that without a critical mass of
participating members, a community cannot exist.
How members interact with their community is ultimately defined by the
available interaction mechanisms and the particular needs of the member. There are
several reasons why people may choose to become part of a FOSS community. The
bulk of the literature on this subject has focused on the motivation of developers
(Hann et al 2004; Hertel et al 2003; Lakhani & Wolf 2003; Scacchi et al 2005;
Schofield & Mitra 2004). Suggested reasons include; pragmatic reasons for needing
specific software functionality, enjoyment of software development as a hobby,
educational benefits, feelings of belonging to a community and/or to a large scale
movement, the need for recognition, self-gratification from a sense of achievement,
and career advancement though skill acquisition. Although the above work gives
some insight into the reasons members have for being involved in community-based
FOSS development, it does not provide a whole picture of motivation in FOSS
communities beyond software development, nor how members’ perception of the
community defines their participation
2 Research Method
The data collected for this research used a predominantly quantitative on-line survey
method. Reaching members of FOSS communities for data collection is inherently
difficult, for the reasons of intangibility and levels of participation explained above.
The sample set of this research consisted of a particular type of Open Source group
within the UK, the Linux User Groups (LUGs). The term is slightly deceptive as most
of these groups do not only concentrate on the Linux Operating System but on a wide
variety of other Open Source operating systems, application and programs. The
research findings presented in this paper are based on the 145 survey submissions
Participation in Free and Open Source Communities 223
Although the survey was directed at the UK LUGs, it was open for others to
participate. Analysis revealed that of the total number of submissions, approximately
12% came from people who were not part of a FOSS society, club, or user group.
Many of the LUGs are involved in software development in some way, and members
may also be involved in other software development communities. The survey used
dealt with individuals’ experiences of on-line FOSS communities in general, not
specifically the LUGs, and although for some members, experience of a FOSS
community will only be the LUG, others will certainly have a broader experience
including other communities. The survey results demonstrate this, as many members
have referred to other communities in their submissions.
The survey itself dealt with several aspects of FOSS communities and the attitudes
and participation of community members. This paper covers the areas of the survey
that collected data about the specific reasons a member may have for participating, in
terms of the actual activities involved, and how and for what purpose a member
makes use of communities.
3 Research Findings
The basic motivation for anyone making use of an on-line FOSS community is to
perform some function, i.e. to use an on-line tool to achieve a desired action. It is
which functions a member uses and why they use them that the initial phase of the
research attempted to discover. This section of the survey collected community
members’ perceptions of what they actually do within FOSS communities and the
pragmatic reasons for participating. Research subjects were presented with several
possible reasons for making use of on-line FOSS communities;
To find out how to perform a task in a software application (Problem solving).
To help other people to use software applications (Providing support).
To suggest alterations or improvements to software programs (Peer review).
To contribute bug fixes or code improvements (Software development).
To meet people or talk to people with similar interests (Social exchange).
The survey question was designed to allow members to select more than one reason
or to specify one or more of their own. Expressed as the actual number of choices,
figure 1 shows how many members chose the above reasons i.e. 127 members chose
(not exclusively) problem solving to be a reason for participating in a FOSS
community. Figure 2 shows this data presented in percentage form (i.e. 25% of all the
choices submitted by all members were for providing support).
As not all members of FOSS communities are developers, it was expected that the
peer review, and software development factors would be less popular than those
relating to support. In addition to these choices, members also posted other reasons
including: being the leader/manager of a community, lurking (Members may have
224 Andrew Schofield and Professor Grahame S. Cooper
many reasons to lurk perhaps bom out of a simple interest in observing discussion), to
encourage the advocacy of FOSS, to build business relations, to learn industry
standards and trends, and finally, just for fun!
Figure 1: Reasons for Participation
The first phase of the research identified the reasons why community members
participate in FOSS communities, in terms of what activifies they are involved in. The
next phase of the research was to investigate how these community activities are used.
exchange-^^H ^ f c l ^
^ – .
Peer review j
^ ^ ^
Figure 2: Reasons for Participation as a Percentage
Participation in Free and Open Source Communities 225
and to collect self-reflective perceptions of why they are used in the manner to which
mIusuallydo not useforums.
B I read what others have said but
rarely participate myself.
a I sometimes participate but only
when it’s useful for me to do so.
Figure 3: Use of Support Forums
the member refers. This phase of the research was split into
looking exclusively at the community aspects which provide
software, and the second at the aspects revolving around software development.
The members were presented with the following alternative ways of interacting with
FOSS support community forums:
I usually do not use forums.
I read what others have said but rarely participate myself.
I sometimes participate but only when it’s useful for me to do so.
I often participate to help both myself and others.
I often participate primarily to be social.
Many members chose to leave additional comments for this question, almost all of
which stating that they preferred mailing lists to discussion boards. There was some
suggestion that discussion boards were more for beginners, and that they are more
focussed on specific issues as oppose to mailing lists which have more general
coverage. The interface of the majority of discussion boards was also criticised and
listed as another reason for members preferring mailing lists.
Finally, those members with software development experience were asked how they
use FOSS community funcfions for software development. The following possibilities
were given and, as before, members could specify their own alternatives.
I mainly participate just to gei help with my own development work.
fmn a I often
participate to help both
participate primarily to be
• I often
two sections; the first
support for the use of
226 Andrew Schofield and Professor Grahame S. Cooper
• I participate both to receive help myself with my own work and to help others with
I mainly participate to get involved in the development projects of others.
e 1 usually do not use forums.
a 1 read what others have said
but rarely participate myself.
D 1 sometimes participate but
only when it’s useful for me to
D 1 often participate to help both
myself and others.
Figure 4: Reasons for Participation as a Percentage
I mainly participate to be sociable.
The order of the questions in the survey and, the request that the completion of this
question is by developers only, is based on the assumption that all developers are also
users of FOSS software. More specifically this means that both users and developers
will make use of the support forums, but that only developers will make use of the
software development forums. It is acknowledged that in some cases these may be the
m I mainly participate just to
‘”‘^”^^^iY WMf) get help w ith nny ow n
ggjM^’^t^ii>i^i::g development projects
‘> 1 ^ ^ WMmmm B I participate both to
receive help myself w Ith
my ow n w ork an to help
others w Ith theirs
m D I mainly participate to get
Involved in others’
a I mainly participate to be
Figure 5: Use of Development Forums
Participation in Free and Open Source Communities 227
same forums but it is still possible to separate the two activities.
Again the members were also given the opportunity to provide their own answer to
the question in case none of these options were appropriate. For this question
members were asked to choose only one option from the list. Figure 3 shows the
choices made by the members and Figure 4 the results as a percentage.
Figure 5 shows the results of this question being put to the developers and Figure 6
shows the same data in a percentage format.
Other uses specified by the developers were: to use the development forums as a
source of research material, to disseminate software to others, to use FOSS
development activities for personal professional development, and again, just for the
fun of it.
Figure 6: Use of Development Forums as a Percentage
4 Research Analysis
By their very nature, FOSS development and the communities performing it are open
to anyone who wants to get involved at any level. The fact that they are also
facilitated by the Internet means that a community is not usually confined by any
geographical constraints, but rather exists on an international or global scale. It is this
fact that justifies the use of the UK Linux/Open Source User groups as the sample set
for this research. The groups may have members from all over the world and each
member is likely to be involved with a myriad of other diverse communities. The
collection of the data for this research itself is a good example. The request for
m I mainly participate just to get lielp
w itii rrv ow n development
B I participate both to receive help
nryself w ith ny ow n w ork an to
help others w ith theirs
D I mainly participate to get involved
in others’ development projects
a I mainly participate to be sociable
228 Andrew Schofield and Professor Grahame S. Cooper
participation was sent to specific UK groups and resulted in submissions arriving
from many other countries which were not specifically targeted. An acknowledged
potential limitation of the research is that LUGs are perhaps more likely to focus on
support than other kinds of FOSS community. There are some communities that are
almost entirely focused on software development and much less on support. Although
many LUG members are involved in other communities there is no way of proving
that the members reached by this survey are entirely representative of FOSS
community members in general. It may be that communities are far more focused on
software development than has been demonstrated by this research. Furthermore, it is
acknowledged that the data has only been collected from FOSS community members
who are not opposed to filling in surveys. This of course is a potential problem for all
academic research but as a person’s views on surveys are not directly related to their
views on the subject matter, this should not significantly distort the results.
The research has investigated communities that are involved with both support and
development activities and consequently has collected data from the different types of
members. The data has shown that in terms of support, problem solving is the main
reason that members have for using FOSS communities, concurring with the work by
Lakhani & Wolf (2003). Interestingly however, only slightly fewer members chose
providing support as a reason. This suggests two things; firstly that the majority of
FOSS community members, in this type of community, perceive support as being the
primary reason or function of the community. Secondly that members rank getting
help from others, and giving it to others, as equally important. The moral views of
Stallman (1999) therefore may be just as applicable now as they were during the early
years of Free Software . Although it is possible that members who prefer to receive
support rather than give it may be less likely to fill in a survey, the significant number
of members who chose providing support as a reason for participating, shows that this
view is common among FOSS community members. It also shows that those
involved in FOSS are aware and appreciate the importance of sharing and
collaboration in community systems as well as software development.
Members also saw peer review and actual software development as being of equal
importance. Since peer review can be performed by member who may have little or
no knowledge of software development, in the programming sense, this highlights the
importance of the user in the FOSS development process and the close user-developer
relationship that exists (Scacchi 2005). It also demonstrates that FOSS communities
are highly involved in the development of software, even when many of the
participating members are not contributing code and may not even be programmers.
These contributions would instead be in the form of software testing, bug reporting
and general suggestions on function and operation (Pavlicek 2000; Moody 2001;
Raymond 1999). If these results are to be considered representative of FOSS
communities in general, the results would suggest that only approximately 50% of
member activities within the community are for reasons of software development.
This supposition is however dependent on the factors of survey participation and
sample set community types.
Participation in Free and Open Source Communities 229
An extremely interesting result was the apparent importance of social exchange
within the communities. 70% of the surveyed members, stated that meeting and
talking to people with similar interests was one of their main reasons for their
participation. This made up 21% of the reasons for member participation (See Figure
2). Sagers’ (2004) and O’Mahony’s (2004) work would seem to fit in with these
findings. However, in specific terms of support and development (See figure 4
through 6), only around 1% of members felt that social factors drove them to use
support or development forums. This suggests that the social activities within the
communities are not confined either to support or development activities but instead
extend to broader social interest.
The second phase of the research, investigating how members use the communities,
has also produced some interesting results and helped to define the different types of
members that make up a community. From this sample set, the majority of members
(36%) use communities for getting support with their software and giving support to
others. Logically this means that many members will login to a FOSS community
website only to help others with their problems, quite possibly with no tangible
benefit to themselves. This correlates with the results of the first phase, in which 25%
of members listed providing support as a reason for participation (See Figure 2). A
slightly smaller number of members stated that they would participate only if it was
useful for them to do so, suggesting that, in terms of support, the two types of
community members are those who perceive giving and taking as being equally
important, and those who require some incentive or personal benefit for them to
participate. Additionally, Zhang & Storck’s (2001) research into “peripheral
members”, supports the research’s finding that approximately 23% of members will
observe the community but rarely participate themselves. This too could be a matter
of incentive but is a very difficult subject to research given the apparent unwillingness
of the members to participate. It is quite possible that there are a great deal more
members that very rarely participate in the sample communities and consequently
were not reached by this survey.
The members of the community involved in development provided a much more
clear-cut set of results. The majority of them (66%) stated that they were involved in
FOSS development communities both to get help with their work, and help others
with theirs, again demonstrating the attitude of collaboration and team work that
exists within FOSS. Only 20% of members said that they participated only to get help
with their own work. This mirrors the findings from the support communities but
indicates that the bi-direction collaborative aspects are more important in actual
software development. Only a very small number of members participated to get
involved in others projects. It is likely that these will be new members, attempting to
get involved with projects for educational purposes.
230 Andrew Schofield and Professor Grahame S. Cooper
The presented research has extracted information about FOSS communities from the
very members that they consist of. It is this unique viewpoint that has revealed the
very interesting inferences that have been taken from the research findings. It has
looked at the ways in which members of a FOSS community perceive the group that
they are in, and has revealed some of the very specific motivational aspects involved.
Although FOSS communities are still often seen as ad-hoc and chaotic, the research
has shown that it is common interest and community relations that bind these
communities together, and allows them to produce both knowledge and software in
such an effective fashion. The research has demonstrated that there is strong sense of
sharing and collaboration within communities that support FOSS development and
use. This manifests itself in two main ways, firstly in the areas of software
development where code, ideas and suggestions are shared and secondly in the
software support area, where information about software use is the object of transfer.
It is this code and knowledge generation and transference between community
members with diverse sets of expertise and backgrounds that allows FOSS
communities to function so well.