Mangione: .Net influences server strategy

The notion of a common data store that spans all of its server products has been something of a Holy Grail at Microsoft Corp. for several years, at least since the days of Windows NT. Microsoft's corporate vice president of SQL Server, Gordon Mangione, sat down at the company's Fusion partner conference to discuss how Microsoft is unifying its data system, tying the .Net servers together, and trying to rekindle the technology that was previously part of HailStorm.

Mangione: The biggest thing is just from an interoperability perspective -- it's the biggest difference that .Net adds into these servers. The protocols and the way you access these things, they're all getting standardized. We may not be there today but the reality is, the way you're going to interoperate, the way you're going to bring it together is all going to be over standard protocols. I think the biggest thing in the 10 years I've been at Microsoft is to finally realize that protocols really do matter. From sort of an internal focus of this, it's the first time we put all the servers on one team at Microsoft and we took my boss, Paul Flessner, and said, "You're responsible for the enterprise server strategy at Microsoft." That's the opportunity for us in making things better than the sum of the parts; getting it so they can fit together, they're managed the same way, there's a consistent way to get support off of them. That's the piece that I think is the biggest difference.

Mangione: Just so we're all clear on the terminology: By .Net I really mean the development environment and how to build Web services and how to bring these applications together. Probably the biggest things that we've done there are the Web services tool kits for Exchange, SQL Server, the e-biz products that really allow you to expose those servers as Web services. The main reason is that tools like VisualStudio.Net know how to connect to these things and know how to bring them together and build applications together. And that's really the message and the technologies that we have today and we've done some work with the 2002 products to bring .Net in closer to those environments, but it's all about bringing these things together in a set of tools for developing applications on top of it.

Mangione: Servers are a funny thing. I mean, servers are probably easier to upgrade than clients just because there's fewer of them. But the lifespan that a server has is very, very long. SQL 2000 has been out for almost two years now and it's probably hitting its stride now. Now is probably the time because the references are there and the people that have deployed it and its peak selling kind of picks up at this point. So in some ways, the irony is that it's really the products now are hitting their stride. They've got a lot of additional capabilities that are built on top of them, but they also have the proof points and the partners are up to speed and the training is there ... Our message for application developers is VisualStudio.Net and SQL Server is the best way to develop .Net applications. And if you ask about where we want people to go and how do they get ready for the next phase, it's build on top of those products today.

Mangione: That's a good question. If you look at Exchange, you've got a hundred million clients on that code base today and you've got to move them forward. There are things we're going to do with that code base and we're going to take it forward. I mean, come on, moving [Exchange] over onto a new storage subsystem -- it's going to take time. If there's one thing we know from a year ago, it's that it's going to take time for us to get there and we're working on Kodiak and we're moving it forward and it's great internal stuff, but it's not like we're going to drop off the face of the earth while we're waiting for those new releases to come out. Getting to a single storage platform and query capability is going to add a ton of capability for customers. Just having a single way to look at that info and query across it is going to be a huge leap forward.

There's a set of tools that work on top of Exchange today and we really encourage folks, if the data is in Exchange that you want to actually program against, you should be programming against Exchange. If you're developing an app that doesn't care about the data in Exchange and you're looking for the best-of-breed tools and the best-of-breed storage, you should be building on top of .Net and SQL Server. That's the best path forward. But if you're caring about mail messaging, collaboration, contact, and calendaring, most of that information is inside Exchange today and those should be the interfaces and tools that you're using to get access to them. We have to do a better job of Web services enabling stuff and building better tools on top of it. But I kind of view that as independent of the big, bold investment of really going to Kodiak. And these two things run side by side and along with each other.

Mangione: 2005. I don't know what the dates are, to be honest. I wouldn't want to speak as to their dates. The thing you have got to do with this stuff is you [have] got to get it right. You [have] got to be able to do all the enterprise messaging and other things. This stuff's like oxygen. [If] messaging doesn't work inside an organization, it grinds to a halt at this point.

Mangione: I think the protocols have to collapse and be the same whether you're client-to-server or client-to-client or peer-to-peer. They've just got to rally around Web services. There's enough industry momentum around this that if you were to try and invent a different protocol for doing things ... I think the whole thing would fall apart. And I think it's one of those things that once the industry adopts it, a whole other set of tools [will] get developed for it, like firewalls, caching proxies, all those types of things. So I think you're swimming upstream to do something different.

Mangione: That's true. That's true. That's true. That's true.

Mangione: I'm not familiar enough about what we're doing with SIP and a lot of the peer-to-peer stuff.

Mangione: A lot of it ends up being, at least internally at Microsoft, tied around work that we're doing in Windows and networking. Frankly, with SIP we don't just want to have instant messaging. We want this to be an infrastructure that people can actually build applications on top of, so it's more about the platform aspects than just being able to sell a service like instant messaging.

Mangione: It's an interesting aspect. I kind of look at BizTalk in regards to how you bring together disparate systems and how you build applications that exchange information with a great set of tools. BizTalk in a peer-to-peer environment -- it's probably not the focus of that team today. They're much more focused on what do I do with these 25 different systems and interoperate between them and what can we do to make partners better able to use that stuff by building connectors for it. From a peer-to-peer perspective, most of what we do peer-to-peer we're going to want to be pervasive. You're going to want to see things like the Common Language Runtime have those capabilities in them and you're going to want them to be extensive so that people can use them. I don't think you want them to think, "Oh, to do peer-to-peer I've got to go buy an enterprise product in order to bring it together to go and work on those spaces." I think if you look at BizTalk it's so much more about how to orchestrate all those systems together.

Mangione: The opportunity is how to make these things better than the sum of the parts. How could Office be the best way that you've used UDDI information? How do you get the type of relationship we have between Outlook and Exchange across the rest of the products. How do I go and build something on SharePoint to collaborate, but I want to use Content Management Server to go and do more professional stuff on the back end? How do I take a simple workflow example that I've started with on SharePoint where I'm approving a document and, well, I really want to orchestrate it with my other systems and I want to plug BizTalk on the back and make those scenarios work better?

Mangione: I think the best business opportunity for us is to think about how to get solutions out of the box in Office around SharePoint, which today is really more [about] how end-users collaborate with info. There may be some APIs on there, but it was mostly an out-of-box scenario. You really want to be able to say, "You really juice up that experience or you really make this a more enterprise-class solution by adding these products on the back end." That's really the types of marriage we want to find between those enterprises. Smart tags being a great example, you want to be able to build smart tags that are in Word documents that take you to your information. You're looking at your live data inside a doc, you want to be able to connect up to your business intelligence systems.

Mangione: We looked at it and we got a couple of interesting pieces of feedback. One is we went to the PDC and we said, "Hey, there's going to be these Hailstorm objects that are going to run on top of Microsoft data centers. You're going to connect up to them." And everyone came back and said, "Nah, it's not what we wanted to hear." They said, "Look, I want to run Web services inside my organization. I want to run Web services that have nothing to do with MSN and the like. What we really want to be able to do is Web-service-enable all of our applications and we want partners and customers to deploy these things independent of having to come back to a centralized service with that information." That's the biggest piece of feedback we got from partners.

Mangione: No, no. And a lot of that good thinking has gone into how we're building Web services on all these other applications. We took a lot of the schema work that they did for contacts and a lot of the schema work that they did for mail messages and said, "This is what we should be doing on top of Exchange." We took a look at the ways that they said, "Here's how you should do generalized data access," and we said, "Geez, that should become the Web services toolkit for SQL Server." We looked at a lot of these things and blended it back into the products themselves because the feedback we got was customers didn't want a centralized service. But one piece that's still there is the Passport piece, because there is still a goal. But you know, I don't have to have 17 different passwords and log-ons for 17 different services if I so choose that I want a federated architecture.

Mangione: A lot of what we've done is taken that thinking, and even some of the folks on those teams have blended them in to do exactly those types of things. Are we calling it Hailstorm going forward? No. We're calling it Web Services and .Net. But [with] a lot of the thinking around schema -- how you get multiple objects, all that stuff -- frankly, there was in some cases overlap with other parts of the company. We were heading down a path of developing multiple ways to access data, multiple ways to store a contact. And that, frankly, didn't make sense. And the message we got loud and clear from developers was, "You already have a bunch of this stuff in other products. You better go figure out how it interoperates." And the right thing we came out with is to say, "We have ways for accessing these things and we should just build it into the products themselves."

Mangione: Absolutely. Absolutely. Absolutely.

Mangione: I don't know if I'd use those words. I'd say we kind of morphed it.

Join the newsletter!

Error: Please check your email address.

More about MicrosoftMSNUDDI

Show Comments

Market Place