Browser-based apps: Don't miss the virtualization forest in the desktop trees

Sometimes we can be so busy concentrating on matters at hand that we miss the bigger picture. Those involved in desktop virtualization have underestimated the significance that browser-delivered applications will have in the future. We have been so busy looking at ways of virtualizing client operating systems and applications that we have not been paying sufficient attention to a shift that will have a very significant effect on what we do in the future.

This is an example of what I call the 'Detroit problem': Way back in the 1960's and 1970's Detroit automakers were aware that Japanese cars were selling in the US but did not take the threat seriously. They rarely saw a Japanese car -- people in Detroit bought and drove American cars. Most people were involved in the Detroit car industry one way or another and bought the local product. In many ways the success of the Japanese vendors was invisible to them. Similarly, those of us primarily concerned with delivering Windows desktops and applications just do not see the changes happening in other areas. Increasing deployment of browser-delivered applications will change some of the fundamental assumptions about desktop virtualization.

The ultimate goal of IT is to deliver applications and data to users. In terms of delivering the applications, browser-delivered applications represent the most significant change for many years. Adoption started around ten years ago with the availability of application servers such as IBM's WebSphere, WebLogic (now Oracle), Microsoft's .NET on IIS and a number of others. Adoption is now significant. For many organizations and software vendors browser-delivered applications have already become the default choice for new applications. There are a number of powerful arguments in favor of browser-delivered applications: