I gave recently a presentation for Capgemini’s Indian Architects Community about delivering high-scalable SOA with Amazon Web Services where I was discussing the potential of Amazon’s Simple Queue Service (SQS) to become the Internet’s Service Bus. Today I stumbled on a project called Gnip (http://www.gnipcentral.com) that wants to “make data portability suck less” by acting as some kind of protocol bridge between several data producers (like Twitter, Flickr, etc.) and consumers. While the service is still in its infancy, not that many big data producers signed up and couple of services like message transformation not present yet, it reminds me to some kind of domain-specific Enterprise Service Bus (ESB). One of the problems we face now with all these Web 2.0 websites that offer APIs and all these data aggregators/clients is the tight coupling between producer and consumer. Service-Oriented Architecture promotes loosely-coupled services and that is exactly what Gnip tries to provide: some kind of message middleware that decouples the tight coupling between Web APIs while providing message transformation (coming soon). If Gnip (or some other service for that matter) succeeds in this by having all the major Web 2.0 companies signed up, they could become the Internet’s Service Bus. Which could be bought by Google so that they can throw in their infrastructure to make it scale and in the meanwhile add some advertisements… ;-) Or of course Amazon by using their Simple Queue Service… Oh, this promises to be an interesting battle, since the company that will own the Internet Service Bus will control pretty much of the Web 2.0 space. Internet Service Bus: the next revolution.