Are the Data Sharing Problems with SDIs Problems of Zombies?
ISBN 978-85-88783-11-9
Authors
1Harvey, F.; 2Coetzee, S.; 3Cooper, A.; 4Iwaniak, A.
1LEIBNIZ INSTITUTE FOR REGIONAL GEOGRAPHY Email: francis.harvey@gmail.com
2UNIVERSITY OF PRETORIA Email: erenacoetzee@gmail.com
3THE COUNCIL FOR SCIENTIFIC AND INDUSTRIAL RESEARCH Email: acooper@csir.co.za
4WROCŁAW UNIVERSITY OF ENVIRONMENTAL AND LIFE SCIENCES Email: adam.iwaniak@gmail.com
Abstract
Spatial data infrastructures (SDIs) have proliferated with the aim of becoming the ubiquitous infrastructure for sharing geographic information but they are neither everywhere, nor guaranteed successes. Ironically, after huge investments SDI too often seem to impede rather than to support free and open sharing. Why is that? We explore in this paper a way of thinking about barriers to SDI success drawing on the popular concept of zombies. We start with looking into the origins of the SDI as a concept from another time - a time before the widespread use of the internet, a time before the web, and a time far before location based systems and the nascent internet of things. If we consider that SDI rests on ideas conceived during the waning of hierarchical modes of information distribution and the first ascendance of the internet and web, then possible reasons for current limitations lurk in pre-internet mapping concepts, including map layers, large centrally-controlled libraries of institutional data and products, and expensive groups of experts who collect, update and publish geospatial data. Sustained through portals, legal protections, mis-directed funding and oligarchic or monopolistic positions, SDIs became a means of extending control over geospatial data and creating other limitations on web-based data sharing. This was not the desire, nor plan of its promoters or original implementers. We suggest in this paper that these two types of limitations are expressions of the zombies lurking in SDI concepts and practices that subtly and surreptitiously exert their influence to constrain data sharing and siphon off resources and programs. As a result, we end up with SDIs that look like SDIs and even act like SDIs, but at the end of the day their zombies sap energy and resources for their own goals and deflect SDIs from their potentials. Why think of these issues in terms of zombies? Zombies occupy an interesting niche in current media and have become a widely used analogy. For example, Labour politicians in the UK refer to do-nothing Tory zombies in parliament; the U.S. Department of Defense prepares for a zombie attack; almost impassive parents struggling with the demands of small children in challenging situations are referred to as zombies. Even philosophers use this term to construct arguments to support propositions against forms of physicalism, such as materialism, behaviorism and functionalism. They call them P-Zombies. There are many different analogies that refer to zombies from which we can start to try and articulate what impedes SDIs. We consider Unix-based operating system zombies as the key reference. In Unix parlance, a zombie is a process without a parent process. A zombie process can hold on to resources that could have been used for other processes. Analogously, SDI zombies take up resources in institutions (long and aimless meetings, unnecessary staff gatherings, unread reports, budgets for duplicate spending, scope creep, meandering projects, cushy deals), with little or nothing of consequence to show. This is a provocative view, but the evidence that SDI activities often result in limitations on sharing information (Harvey & Tulloch, 2006; Harvey, 2003; Nedovic-Budic & Pinto, 2000; Obermeyer, 1995; Masser & Campbell, 1995) suggests that SDI zombies are actually quite widespread. How widespread are zombies in SDI? Provocatively, if you want to see the evidence of zombies, try to read some metadata. A concept with good intent becomes a daunting barricade that often camouflages old institutions with syntactically correct and informative meta-information, but the access to data has not changed and the metadata zombies cunningly depreciate its potential. In a zombie-influenced SDI we end up with situations where people talk about sharing, but actually are interested in controlling and getting data to feed their public kingdoms or private businesses. Impaired data sharing and SDI results. In many cases, super-com
Keywords
SDI; Potentials; Data sharing