Each year, Google offers stipends to students for working with selected open source projects for three months. This year, I’ve attended the program by contributing to Eclipse’s Code Recommenders project, lead by Marcel Bruch, who also mentored me. A detailed description of the results are already given in an earlier post. Since Google Summer of Code officially ended a few days ago, it is now time for a résumé.
Prior to the actual program, students have to contact open source organizations, which successfully have applied to Google (175 this year), and to agree on potential projects, as well as to look for potential mentors. Eventually, the student will register his project proposal with a website set up by Google, which allows the organization’s mentors to rank the proposals related to their organization. Finally, Google will decide on how many projects each organization will be supported with and the highest ranked proposals will be accepted accordingly (1115 passed the barrier). Then, the fun begins with 4 months of coding, supervised by the associated mentor.
Since I had already worked with Code Recommenders for a university hands-on training before, there wasn’t much preparation to be done, we could literally start working from the first day on. Also, since the program process is obviously mature (it’s in place since 2005), I’ll rather comment on the whole concept of financially motivating students to work with open source development, which usually is based on non-financial motivation. The general pro’s and con’s of open source in contrast to proprietary software shall not be discussed now. Nevertheless, once someone decides to take part, either by starting an own project or joining an existent one, there are several obstacles: either one has to market the project to attract a community or one must be integrated in the developer team, in both cases usually hundreds of hours of work are required to create successful results.
Having a financial motivation, which at the same time also is an obligation, the participant is more likely to stick with the project through good and bad. The 4 months of sponsored work should typically be enough to surpass the mentioned barriers and to comprehensively familiarize oneself with the project, making further engagement unproblematic and also much likely (since the student is assumed to have a general interest in the project he applied to). As a result, a relatively small investment tends to benefit all participating parties.
Speaking for myself, this concept proofs to be quite effective. The 4 months gave me the opportunity to be part of a large project, which isn’t in any way behind traditional, commercial endeavors, which offered experience, insight and inspiration for further projects. As a conclusion, I think that hybrids of commercial and voluntary work will have a promising future in educating students, attracting developers to public domain projects and creating products also commercially operating enterprises benefit from without holding exclusive rights.
Finally, I would like to thank Google and Eclipse very much for offering this great opportunity, as well as my mentor, Marcel Bruch, and the Code Recommenders co-lead Johannes Lerch, for making my participating possible and offering great support and frequent feedback on my work!










