Improving time-to-market through globally distributed software development.
Reductions in development time and cost are important factors for software engineering.
Such reductions not only enable companies to sell products earlier, faster and cheaper but
in some cases, are necessary for the survival of an organisation. While many ways have
been suggested to reduce development time and costs, the complex nature of software
development remains a challenge.
This research investigates the opportunities to reduce time-to-market through "around
the clock" software development. It focuses on a particular work pattern, SeSE
(sequential collaborative software engineering), in which a task is passed at the end of a
working day from one developer to another across time zones. The research identifies the
related contextual factors and associated overheads and presents a model of the
relationships between development time and these contextual factors and overheads.
As part of the evaluation of this work, the thesis presents an exploratory empirical study
that has been carried out to validate the feasibility of this work pattern and to obtain
values for some overheads. In addition it presents an implementation of the model that
demonstrates one application of the model and facilitates its verification. The thesis
concludes that software development across time zones is a feasible work pattern and can
potentially be used to reduce development time. However several related issues have to
be addressed before this work pattern can be widely adopted and become normal practice.