Building the P2Pvalue software platform

One of the key parts of the P2Pvalue project is “Integrating the functionalities of online social networking services and collaborative software in a privacy-aware platform based on a decentralised architecture”.

At the beginning of the project the software team were working on building this platform on top of the established Kune free/open source distributed and collaborative social network software. Kune has been in development since 2007, mainly by Vicente J. Ruiz Jurado together with other contributors, and under the umbrella of the Comunes nonprofit. The current result is a feature–rich, stable, free/open source software platform translated into 20 languages and hosting over 1100 groups.

However, we began to question whether we should be developing something that was essentially feature-driven, as Kune is. We decided to research user needs and concluded that we should provide CBPP communities with a design with specific and clear added value that arises directly from the needs they express.

Kune uses the Apache Wave technology as its foundation. Apache Wave provides a framework to develop collaborative software with some nice features such as content federation, real time edition and gadgets. However, the effort to adapt Kune to the new value proposition was considered to be harder than building the tool directly on top of Apache Wave. Consequently the decision was taken to do just that.

This will allow us to continue to take advantage of the protocols and potential of Wave; it will support the privacy-by-design approach right away; and this line is considered feasible because the team has spent the last months studying Wave’s complexity and stabilizing its code. This way, we will be able to follow the bottom-up design process (we continuously ask the users about each new rapid prototype, asking which features they feel missing, and implementing those), instead of our previous more top-down approach (in which we decide that the users want “this” so we implement the corresponding features).

The first step to get user feedback and find user needs was to run a couple of workshops in Tabacalera, a well-known self-organized social center of Madrid. The two workshops were attended by around 20 members of several CBPP communities. The results of the workshops were published here on the P2Pvalue blog.

Three value propositions for the new software were defined: an information and coordination mobile tool; a dissemination tool based on community and individual maps; and a tool for communities interaction (as explained better in the slides embedded in the mentioned blog post).

After an evaluation from the UCM team of the needs of the P2Pvalue project, we decided to focus on the information and coordination mobile tool, and specifically on task management and participation visibility.

This topic is fully aligned with P2Pvalue objectives; if we are interested in how much value each participant is generating, we need to know which tasks she completed. In any case, this is just a starting point in the user-centered design. The goal is to provide value quickly and build the tool from user feedback.

To start the process of user-centric design, we carried out a set of interviews to learn more about task management and participation visibility among CBPP participants. Around ten interviews were done, asking for aspects such as where tasks are decided, do people give thanks to others, how people assign tasks to themselves etc.

These interviews showed 4 types of tasks: decided in a meeting (face-to-face or virtual), unexpected events, self-promoted and periodical tasks. We decided that the most interesting for us right now were the tasks that are decided in a face-to-face or virtual meeting.

The next step was the building of a user journey, which describes how a participant of a CBPP has an idea, how she introduces it in a meeting, where appropriate tasks are discussed and assigned, and how other users actually do the tasks and report that the job is done. We modified the user journey with how our tool could help these users in this process.

Once we had our desired user journey, we described the minimum viable product (MVP) so we could build a minimum prototype, show it to the users and learn quickly from their feedback.

The minimum prototype consisted of a task list with assignees in a mobile friendly format, that could be used in a CBPP meeting, where participants typically have their mobiles at hand. This prototype was designed by a professional designer and follows the P2Pvalue design guidelines.

The prototype is available at https://prototype.p2pvalue.eu

The first interactions showed many usability issues, as well as several requested features such as deadlines, task alarms or personal tasks.

We are following an iterative process of prototype design, building the new design, showing the new prototype to CBPP participants, gathering feedback from them and discussing the new prototype design. We have estimated that each cycle (sprint) will take us 8 days. We are using scrum to manage the process.