The following hypotheses regarding the software development practices of this end-user community were posed as open questions to 15 computer musicians who agreed to participate in an interview. Section V presents the interview recruitment strategy in detail. Relative to the general population of software developers we hypothesize that:
1) Computer musicians make less commits.
Responses: 7 computer musicians agreed with this hypothesis,
1 disagreed, and 7 were uncertain.
Rationale: Advocates of the hypothesis argued that:
◦ “Computer musicians are used to working solo ... there’s less incentive to keep source control repositories up to date.”
◦ “The community [of computer musicians] resembles a musical community more than a development community. The culture of sharing ideas is different.”
◦ “It is really hard to differentiate minor changes from actual, structural changes in your code. After a while, it gets pedantic to commit changes like ‘Changed parameter X so it sounds more like a guitar’.”
The computer musician who disagreed argued that music
projects are no different than general opensource projects. 2) Computer musicians commit more on the weekend. Responses: 4 computer musicians agreed with this hypothesis,
2 disagreed, and 9 were uncertain.
Rationale: Advocates of the hypothesis argued that “many
others have day-jobs unrelated to computer music, meaning their projects are more hobbyist in nature, which may mean the weekend is the only opportunity they have to make solid contributions”. Those that disagreed argued that the number of weekend contributions would differ “only if computer music is your hobby”.
3) Computer musicians make commits less frequently. Responses: 5 computer musicians agreed with this hypothesis,
1 disagreed, and 9 were uncertain.
Rationale: Computer musicians advocating this hypothesis
argued that since computer music software development is often a hobby, commits to repositories are less frequent: “Computer music does tend to be a field in which our output is not particularly commercial (making it a free-time activity for a lot of us).” The computer musician who disagreed argued that music-oriented projects are no different than general opensource software projects.
4) Computer musicians report less issues (bugs).
Responses: 7 computer musicians agreed with this hypothesis,
3 disagreed, and 5 were uncertain.
Rationale: Advocates of the hypothesis argued that:
◦ “It is hard to describe some bugs in words. How do we solve ‘my synthesizer sounds too bright ... ?’”
◦ “Many work solo, and the incentive to learn and update bug trackers is not so urgent.”
◦ “So many computer music people are not initiated into the world of software development. It is a kind of amateurish community when it comes to technical stuff (notable exceptions abound, of course).”
Those that disagreed argued that music projects have similar
numbers of bugs to report as general opensource projects. 5) Computer musicians’ repositories have less unique con-
tributing authors.
Responses: 8 computer musicians agreed with this hypothesis,
1 disagreed, and 6 were uncertain.
Rationale: Advocates of the hypothesis argued that:
◦ “Music projects are generally going to be more creative, and they might only reflect the vision of one developer.”
◦ “The people who slave away at this wonderful software often work alone. Such is often the case, I am afraid,
for audio people in general.”
The computer musician who disagreed argued that there exists large-scale music projects with several contributors.