Spent the day at the SpringSource S2G forum, basically a one-day conference by Spring Source. It was a good experience for me - a good fit for things that I’ve got going on or am thinking about these days - and helped clarify for me what the SpringSource guys are trying to do with the proliferation of products I’ve complained about in the past.
In short, Spring’s focus is on improving developer productivity, and to enable portability of applications. The appeal of the original Spring Framework was that it was much simpler and more productive than EJB, and Spring’s architecture is designed from the ground up to let you plug in alternative implementations, whether it’s the various bits of the presentation layer, persistence layer, transactions, web services, you name it. Roo and Grails are two different approaches, with different use cases, to bring the kind of productivity improvements seen in Rails and similar newcomers to the development platform scene to the Java platform.
STS is really the centrepiece of Spring’s efforts to improve developer productivity, both by providing a pre-packaged set of Eclipse plugins to work with Spring applications, and also as a way to tie together the various parts of the Spring portfolio, such as Grails and Roo, and the various operations-side pieces like TC Server, Hyperic, and even the cloud platforms Spring is working on. The presenters at S2G all used STS, which was an effective way of showing how to get the best out of it and the specific technologies being demonstrated.
Rod Johnson’s keynote was heavily focused on the cloud, in the wake of the recent announcement of VMforce and discussion of their cloud strategy. Johnson explained how cloud fits into Spring’s strategic focus, which goes back to improving developer productivity, and portability of applications. It really does make sense for Spring to make a push into the cloud, since otherwise the Java platform will be left in the dust by Azure in the enterprise space.
The cloud, and PaaS in particular, is about making life easier for developers, and I know that I am certainly in the market for a way to move my J2EE applications into the cloud. At present, the only real option is to roll your own PaaS on top of an IaaS like EC2, which is a lot of work.
So Spring Source’s Cloud Foundry, VMforce, and the other offerings that seem to be in the pipeline will really hit the spot. And they do more than just offering an Azure equivalent, but true to the Spring philosophy, they give us alternatives, whether it’s different services to host on, or even the option to deploy a PaaS in our own data centre. So we’ll have portability, even the capability to deploy our applications across multiple cloud providers simultaneously. I’m sure that the market for helping enterprises create private clouds will be big, even if it’s not “pure” Cloud.
There were a number of other points Johnson made about Cloud that really hit the mark with me. I’m definitely keeping an eye on this space.