I’m at QCon London and covering it on Twitter as wi-fi and battery power allows. The focus of the conference is enterprise development, and it spans Java and .NET, SOAP and REST, but with a bias towards Agile methodology.
I was here last year, and my observation is that last year there was considerable angst about the idea that the SOAP stack was failing to deliver and that new stacks based on REST were the thing to do.
This year this same statement feels more widely accepted and people are moving on, based on that assumption.
One delegate stuck his hand up and asked, “is this a fad, or will it last”? Sounds like a cry of pain from someone who invested heavily in something now condemned. History tells that in five or ten years time we may well be equally enthusiastic about some new thing and observing all the weaknesses in the bad stuff we used to do.
Still, these trends don’t form overnight. We have to choose what looks like the best model at the time.
I’m typing this during a session on test-driven development. Personally I think TDD is a more profound improvement in the software development process than REST vs SOAP; the detail of TDD will continue to evolve and be debated; but its basic ideas are hard to challenge.
As an aside, this is one factor in why ASP.NET MVC is so important for developers on Microsoft’s web platform. Reason: it’s testable.