Adam Bosworth posts this compelling and potentially divisive question on his blogsite. That is, our current generation of commercial databases don't cut the mustard. (And perhaps open source isn't there yet, either.)
Databases don't scale as they should, Bosworth argues. We're all spoiled by Google's indexing and millisecond search capabilities, something that today's database products don't give us. Databases need more flexibility to change; we need dynamic schema so that as the business model/description of goods or services changes and evolves, this evolution can be handled seamlessly in a system running 24 by 7, 365 days a year.
So, does it really matter how good our databases are?
Ultimately, the appeal and value proposition of XML comes from the fact that data can be stored and accessed by any system at any time across the network. We depend on our databases as the key enabler of our SOAs.
But, another take is that as SOAs develop, get more robust, and gain more intelligence, much of the functionality we relied on in databases will be moved to the middleware layer. The database itself will be more of a commoditized repository. A few months back, I had the opportunity to speak with Jeff Jones, product manager for DB2, on this very topic. He says that IBM's DB2 group and Websphere Application Server group are working closer and closer together. No longer can any single piece of the technology stack dominate that stack. "The emphasis is shifting away from the parts and more to the whole," Jones says. "Customers are taking us there, regulatory pressures are taking us there, and on-demand computing is taking us there."