After the 2009 general elections in India a controversy started about the electronic voting machines (EVM) that are used nationwide since 2004. Political parties, activists, and academics raised suspicion that the machines might have been manipulated to alter the election outcome. There is no proof that EVMs have been manipulated in any of the past elections, however, concerned people claim that the risk is there. This paper takes a closer look at the Indian voting technology and the discussions around alleged security holes. The authors take a closer look at this particular controversy. Additionally we want to provide the reader with information about the Indian electronic voting system more generally. This includes reasons to change from the earlier paper ballot system and design challenges for EVM in the Indian context. We are writing within the frame of a theoretical model called Social Construction of Technology (SCOT), developed by Wiebe Bijker and Trevor Pinch (1987). Along the lines of this model we argue that after the EVM has been adopted in India, different "relevant social groups" interpreted the EVM in diverse ways. From the social constructivist perspective we argue there has been not just one but at least three different EVMs. With time the "interpretative flexibility" diminished and "relevant social groups" more or less agreed on one interpretation of the EVM. The EVM has "stabilized" and the controversy has been closed basically. We show the SCOT model to be helpful for structuring the controversy in a fruitful manner. The research questions adressed here are: How did the ECI and EVM manufacturers react to allegations made by political parties, VeTA, and voting security researchers that EVMs are vulnerable to manipulation? How was the election practice affected?