Relationality in geoIT software development: How data structures and organization perform together

Abstract:

Constructivism in geo-information science has emphasized what happens to geo-information technologies (geoIT) after the design stage, when systems and applications are used in real life. Current constructivist views, however, have focused less on other aspects such as software development practices. This paper adopts a similar constructivist epistemology, but looks at how geoIT and people are entangled in the development stages. We discuss the case of the migration of GIS software to Free and Open Source license. This case provides clear empirical evidence of the entanglement of humans and artifacts during the development of GIS technologies. Through an analysis of archived material (such as mailing lists), and of the software code, the paper describes how the integration of a new software (the library Fast Fourier Transformation in the West) was hindered by the different data structures of the original GIS and the new software. The case study we propose shows how actual software development practices may contrast with the well-established rhetoric of technical efficiency of the algorithms. In addition this choice also illustrates the organizational aspects of developing GIS and the different weights that are given to computational resources and organizational resources. © 2010 Elsevier Ltd.