Objects in the cloud may be closer than they appear towards a taxonomy of web-based software

The model of computation in which software is downloaded and/or run dynamically from the network is commonly referred to as cloud computing. Historically, cloud computing has implied the use of thin clients, i.e., an approach in which the majority of computation is performed on the server side. However, in recent years the landscape of cloud computing has become quite diverse. This is partly because client computers and devices in cloud-based systems have become richer and far more powerful, and partly because the capabilities of the web browser as an application platform have increased substantially. In fact, with the upcoming HTML5 and WebGL standards, we anticipate client computers and devices to take a much more encompassing role in cloud computing. Additionally, the rapidly increasing use of mobile devices will have a dramatic impact on the use of cloud-based systems. In this paper we analyze the different flavors of cloud computing, with an attempt to classify web-based computing systems and to facilitate discussion and reasoning about such systems in general.