SOA, BizTalk, .NET, Visual Studio, and "Oslo"

The week of November 5, I attended Microsoft’s fifth annual Service Oriented Architecture (SOA) and Business Process Conference and learned about a new assortment of technologies code-named Oslo. According to Microsoft, this “multiyear, multi-product effort…to build on the model-driven and service-enabled principles of Microsoft Dynamic IT” will “further Microsoft software-plus-services efforts by providing extensions to the application platform to help developers bridge between on-premise and off-premise projects.”

No, that’s not a string of gibberish from some gearhead version of Alice in Technoland. It’s the sound of Microsoft reinventing itself, rejecting its legacy as a desktop company that became a server company, too. Today, Microsoft wants you to know it’s a “platform” company, and Oslo bolsters the platform vision. What that means is that Microsoft has lots of products that do great stuff, but the more important point is that those products all work together to let you do even cooler stuff. According to General Manager of Microsoft’s Connected Systems Division (CSD), Steve Guggenheim, the company wants us customers to look at its products in terms of how “the releases enhance the different capabilities in the platform.”

The term “platform” is one that you hear everywhere at Microsoft these days, and this ubiquitous word means various things in different contexts. For example, SQL Server is no longer a database; it’s a “data platform” and a “BI platform.” Exchange Server is an element of the Unified Communications (UC) “platform.” Forefront is the assemblage for a security “platform.”

However, Guggenheim explained, from his group’s perspective the meaning of “platform” is clear: “I think about the ‘application platform,’ which is what you use to build and run and manage applications. For a long time, ‘platform’ for Microsoft meant desktop—Windows and Office. In the last ten years, we’ve built the server platform side. Now you get into the other side of application development, which is the business application side and the backend side.”

 To compete with platform companies such as IBM and Oracle, Microsoft has to move beyond selling individual products. Guggenheim said, “There’s a natural tension between providing a best-of-breed product—take SQL Server, for example, as being a best-of-breed database—and being part of an overall platform. So we’ve traditionally landed a little heavily on the individual products, as opposed to how these things work together. But at Oracle, IBM, SAP, where you’re competing at the platform level, they do a pretty good job of talking about how all the pieces relate. We do a better job of developing, or building, products that work well together, but we haven’t been doing the articulation as well. As you build more pieces of a platform, people want to know how they’re related. A lot of feedback we get from customers is they’re happy that the stuff actually works together well.”

What’s in Oslo?

OK. So what does all that “platform” talk have to do with this new group of technologies code-named Oslo? The idea is to strengthen Guggenheim’s application platform and make it more competitive. According to Burley Kawasaki, a director in Guggeheim’s CSD team, the Oslo technologies will be delivered across five areas that Microsoft considers part of the application platform:

Kawasaki said that Oslo’s two key areas of investment, BizTalk Server and BizTalk Services, would support the SOA platform, serving as “the anchor for our SOA investments.” He continued, “We’re going to be making additional investments to make it easier to deploy services that aren’t just on premise but stretch across organizational boundaries and into the cloud..”

Further, he explained, “Oslo can be the anchor for taking model-driven development and making it a mainstream part of app development.”

The problem with model-driven application development, Kawasaki continued, is the lack of a common model for business, architects, and IT: “People have been looking at using model-driven tools in very specific domains: Modeling tools for the business might model business processes or requirements or policy. You have a different set of modeling tools for architects, who might model schemas or service contracts. And you probably have a different set of modeling tools for the IT professional, who might need a deployment model or a health model. You have all these models, everyone using their own tools, not connected. When you start trying to build composite applications that pull together bits and pieces across all these, it becomes very difficult to get a whole view of the app. That’s one of the things we’re addressing through Oslo: We need models to not be a description of the application, but the model needs to become the application, and vice versa.”

Who, What, When

This is all very interesting, but also very unreal at the moment. I’ve written before that I see the logic and value to Microsoft of focusing on “platforms,” but it’s hard to know whether any of this will actually happen and whether customers will really jump on board. Who is demanding these technologies? According to Microsoft’s “Fact Sheet” (hmmm… facts about non-existent technologies?): “Traditionally, composite application development has been highly complex and costly—and thus only available to the very largest of enterprise customers. We see ‘Oslo’ as having benefits for companies of all sizes.” However, the “Fact Sheet” doesn’t say what those benefits will be.

 What will be the form factor for Oslo? According to the Fact sheet: “It is still too early to talk about any specific packaging, branding or licensing details.  However, ‘Oslo’ products and technologies will surface across multiple Microsoft product lines.  BizTalk Server, Visual Studio, System Center and Office System will likely continue to be separate product lines, just as they are today, with their own SKUs and release schedules.  However, as ‘Oslo’ products and technologies become available, we expect them to be used in these product lines to add new capabilities across the platform.  Customers with Software Assurance rights will be able to take advantage of these capabilities under their existing contracts.”

When will we see something from this Oslo initiative? The Fact Sheet says, “Because ‘Oslo’ technologies will ship across multiple products, there will in fact be multiple CTPs and betas that deliver various aspects of our R&D investments.  We plan to have at least one major CTP of ‘Oslo’ technologies in 2008. We currently have a number of customers and partners contributing to the direction and development of ’Oslo’ internally, and are committed to driving and prioritizing our development efforts based upon customer and partner feedback; they will help inform us as to the exact timing of our broader CTP and beta efforts.”

The Road to Oslo
I’ve been around long enough to remember lots of Microsoft’s grand visions and initiatives, so I can’t help thinking of that other city-named vision that was code named Cairo. The road to Cairo went nowhere—well, actually, it went all over the place and bits and pieces of that vision have appeared in the past several generations of Microsoft technology. Maybe that’s why Oslo isn’t positioned as a product, but as a bunch of things that will be integrated with other things. I guess it all goes back to Guggenheim’s mission of “enhancing the platform.”

Whew! That’s a Long Blog Post!

Well, if you made it through that long stream-of-consciousness ramble about Oslo, I hope you’ll let me know what you think of all this platform stuff and this odd collection of technologies. Does anybody care?