Code forges are third party software repositories that also provide various tools and facilities for distributed software development teams to use, including source code control systems, mailing lists and communication forums, bug tracking systems, web hosting space, and so on. The main contributions of this paper are to present some new data sets relating to the technology adoption lifecycles of a group of six free, libre, and open source software (FLOSS) code forges, and to compare the lifecycles of the forges to each other and to the model presented by classical Diffusion of Innovation (DoI) theory. We find that the observed adoption patterns of code forges rarely follow the DoI model, especially as larger code forges are beset by spam and abuse. The only forge exhibiting a DoI-like lifecycle was a smaller, community-managed, special-purpose forge whose demise was planned in advance. The results of this study will be useful in explaining adoption trajectories, both to practitioners building collaborative FLOSS ecosystems and to researchers who study the evolution and adoption of socio-technical systems.
[1]
Megan Squire.
Data Sets: The Circle of Life in Ruby Hosting, 2003-2015
,
2016,
2016 IEEE/ACM 13th Working Conference on Mining Software Repositories (MSR).
[2]
Yongqin Gao,et al.
A Research Collaboratory for Open Source Software Research
,
2007,
First International Workshop on Emerging Trends in FLOSS Research and Development (FLOSS'07: ICSE Workshops 2007).
[3]
Georgios Gousios,et al.
The GHTorent dataset and tool suite
,
2013,
2013 10th Working Conference on Mining Software Repositories (MSR).
[4]
Kevin Crowston,et al.
FLOSSmole: A Collaborative Repository for FLOSS Research Data and Analyses
,
2006,
Int. J. Inf. Technol. Web Eng..