AA: The Algorithmic Autoregulation (Distributed Software Development) Methodology (WSL 2013)

Authors
Renato Fabbri (USP/LabMacambira)
Ricardo Fabbri (UERJ/LabMacambira)
Vilson Vieira (USP/LabMacambira)
Alexandre Negrão (LabMacambira)
Lucas Zambianchi (LabMacambira)
Marcos Mendonça (LabMacambira)
Daniel Penalva (UNESP/LabMacambira)
Danilo Shiga (LabMacambira)

Abstract
We present a new self-regulating methodology for coordinating distributed team work called Algorithmic Autoregulation (AA), based on recent social networking concepts and individual merit. Team members take on an egalitarian role, and stay voluntarily logged into so-called AA sessions for part of their time (e.g. 2 hours per day), during which they create periodical logs — short text sentences — they wish to share about their activity with the team. These logs are publicly aggregated in a Website and are peer-validated after the end of a session, as in code review. A short screencast is ideally recorded at the end of each session to make AA logs more understandable. This methodology has shown to be well-suited for increasing the efficiency of distributed teams working on what is called Global Software Development (GSD), as observed in our experience in actual real-world situations. This efficiency boost is mainly achieved through 1) built-in asyncrhonous on-demand communication, documentation of work products and processes, and 2) reduced need for central management, meetings or time-consuming reports. Hence, the AA methodology legitimizes and facilitates the activities of a distributed software team. It thus enables other entities to have a solid means to fund these activities, allowing for new and concrete business models to emerge for very distributed software development. AA has been proposed, at its core, as a way of sustaining self-replicating hacker initiatives. These claims are discussed in a real case-study of running a distributed free software hacker team called Lab Macambira.

Download paper (PDF)

Share

Metrics

Views:

Downloads:

Citations: by Google Scholar