I've had a hard time understanding what Microsoft's Oslo would offer companies -- and I even did an e-mail Q&A with Microsoft's Burley Kawasaki, director of product management in the Connected Systems Division, about it earlier this year.
I'd read everywhere that Oslo was Microsoft's take on service-oriented architecture.
"Finally," everyone said, "Microsoft was getting on the SOA bandwagon and showing us what it had to offer."
But that wasn't what I heard from Kawasaki. And the more I read about what Microsoft has to say about Oslo, the more I began to suspect that it really didn't have that much to do with building a SOA -- at least not in the same way other so-called SOA products promised.
My real turning point came after reading this eWEEK article. Steven Martin, senior director of product management in Microsoft's Connected Systems Division, spoke about Oslo at the Microsoft Worldwide Partner Conference.
He focused -- just as Kawasaki did, to some extent -- on software modeling. Martin talked about Oslo as a "core modeling technology" that could help with cloud computing. The article notes he did talk some about SOA, but mostly in terms of how SOA, Web services, cloud computing and modeling are all "vectoring together."
If I read it correctly, it seems Microsoft sees modeling as the glue that can put these pieces -- services, if you will -- together and hold them in place. And Oslo is Microsoft's brand name for superglue. Said Martin:
Hardware virtualization and software virtualization are two distinct things. That's where modeling comes in. Writing applications that are an aggregation of services is hard, and managing them is hard. Modeling will be the way out of jail for building these huge applications that people are building in the cloud. And that's what we're doing in Oslo. Modeling technology is not only important now, it's critical.
No doubt, if people at Microsoft bother to read this blog, they're thumping their foreheads with a shout of "Finally, she gets it" -- or possibly, "Man, is she dumb." After all, Kawasaki tried to make this point back in January when I asked him whether Oslo was SOA solution or a modeling tool:
Oslo addresses both modeling and services -- these are the two big investment areas targeted by our Oslo investments. ... A world-class SOA platform is necessary to build applications within and across organizations, but it is not sufficient by itself. Building applications can be made much simpler. The second element of our strategy is to make modeling a mainstream part of application development.
But I plead not guilty to blatant obtuseness, because I think this is the clearest explanation of Oslo offered so far by Microsoft. Plus, as Martin points out, Microsoft is fighting an uphill battle on this one because it views modeling much differently than most people. In some ways, Microsoft is redefining "modeling" or at least reasserting its definition as broader than "application workflow." As Martin explains it, Microsoft wants modeling to mean:
- Allowing more people -- including business users -- to participate in application design.
- Allowing developers to write applications at a much higher level.
- Developing models that are accessible and pertinent across the enterprise, rather than models that are created in silos.
- Real-time models, rather than models that are printed out and immediately become obsolete when they're handed off to the developer.
- And, perhaps most importantly, models that are executable because they're managing or calling the services. In essence, Microsoft wants you to be able to build the model and have an application, rather than building a model and then building an application that vaguely follows it.
It's a big vision, and -- I'm certainly no developer -- but it seems to me that SOA would play a key role in creating it. As I understand it, Oslo isn't to help you get to SOA; Oslo is what will help you get to SOA and then leverage it into something really powerful.