Developing Software with Open Source Software Components - Finding Source Code on the Web for Remix and Reuse - page 171

component selection activities. The only requirement for companies to participate
was that they had undergone a finished project that implied OSS component selec-
tion. Organizations' details are given in Table 9.1 . Some respondents came from the
same organization, but worked on different projects as detailed by the fourth column
of the table.
Nineteen respondents from 17 European organizations from Spain, Norway and
Luxembourg participated in the study. These organizations included: software con-
sultancy companies (SCC) that perform software development tasks for different
clients as their primary business; IT departments (ITD) in public or tertiary orga-
nizations that usually perform or outsource some software development task for
covering the internal demands of the organization; to a software house (SH) that
develops and commercializes specific proprietary solutions; and one organization
that provides expert support for selecting software (ESSS) solutions based on their
clients' requirements. However, this organization does not perform any software
development tasks.
Each respondent was asked to talk about a single finished project that he/she
was familiar with, and a single component used in that project. Interviews were
mainly performed in the mother tongue of the respondents and face-to-face in their
working place, by one or two researchers of the team. Interviews lasted around 1 h
each and were recorded on paper and tape for subsequent analysis. The project and
the component(s) were chosen by the respondent without any intervention from us.
The resulting set of projects was diverse, and used a variety of components that
ranged from libraries and APIs to more complex solutions.
Interviews were prepared for analysis by a manual transcription of audio records
to text documents, and were finally translated to English so that the whole research
team could equally assess and discuss the data. We used “content analysis” [ 30 ]asa
basis for performing the assessment of the collected data, and generating categories
by grouping sentences of phrases that described the same idea, action or property.
9.4 Results
This section presents the results of the study. They are grouped in two subsections
according to the two research questions introduced above. Results are described in
terms of the categories or codes generated from the data analysis. Interpretation and
discussion of the findings according to the research question are tackled in Sect. 9.5 .