UOC experts find that open-source software projects do not operate democratically

Two members from this research group have analysed the 25 most popular projects on GitHub, the online platform par excellence that hosts almost all the open-source projects, currently totalling more than 10 million. It is the first time that researchers have analysed the internal participation mechanisms for these types of project.

The researchers have found that, out of the 25 projects studied, only one clearly explains how the project is governed (who decides what and when) and another seven give brief indications, while the rest, which make up 68% of the projects analysed, give no information whatsoever. And those that do say something do not follow democratic practices. So, in actual fact, none of the 25 projects analysed abide by any form of democratic process in their functioning.

Frustrated users

According to the researchers, most of the software projects that criticize how our society operates are run by a small group of people who may take more or less notice of users' requests (regarding errors to be corrected, new functions to be added/modified, new components to be integrated) but do not feel any obligation whatsoever to act on them or, at the very least, explain how the decision whether or not to listen to them is made. In fact, it is not uncommon to hear expressions such as "we have to protect users from themselves" or "they don't know enough about the subject to give an opinion".  Even worse, some of these projects, even some that are quite well known such as WordPress or Linux, are run by "benevolent lifetime dictators" who may make totally unilateral decisions at any time without taking into account the community's opinion.

According to Jordi Cabot, director of the SOM and ICREA Research Professor at the UOC, “this is clearly detrimental to and frustrating for users, who are unable to have any direct influence on the software's future while they are able to influence the evolution of their social environment, and not knowing about laws, for example, does not detract from their right to say what they think and take part in binding voting processes." Cabot acknowledges that "the software has had a very positive influence on democratic progress in our society, for example, by enabling the emergence of electronic democracy. Perhaps the time has come to bring democracy to the software itself".

Tools for transparency

To change this situation, the SOM Research Lab is developing a series of tools and recommendations to transform software development processes into democratic, transparent processes where each project can explicitly define the decision rules. Once they have been defined, these rules can be monitored and automated to produce a software that truly answers the community's needs. They hope to start getting concrete results by mid-2016.

The SOM Research Lab started a general line of research a couple of years ago to analyse the open-source world within the framework of a medium-term research project that, in forthcoming years, will concentrate particularly on studying democratic tools in these environments. They plan to apply for EU funding so that they can study participative software development in greater detail.