Windows chief talks '7'
In an exclusive interview, Steven Sinofsky offers up a few details on the new operating system and the rationale for why he is not saying more publicly.
REDMOND, Wash.--Since taking over the Windows development reins from Jim Allchin, Steven Sinofsky has chosen to fall almost completely off the public radar.
It's not that he hasn't been busy getting Vista Service Pack 1 out the door and starting work on Windows 7. It's just that Sinofsky doesn't want to talk about products until they are well along in their development. Last year, Sinofsky penned a blog to his Windows unit co-workers, explaining his public silence and urging them to follow his lead.
"I know many folks think that this type of corporate 'clamp down' on disclosure is 'old school' and that in the age of corporate transparency we should be open all the time," Sinofsky wrote. "Corporations are not really transparent. Corporations are translucent. All organizations have things that are visible and things that are not."
Well, Sinofsky is breaking his public silence, slightly, to offer a few important details about 7 (he reiterated that it is coming by January 2010) and to explain why he is saying so little publicly.
In an exclusive interview with CNET News.com last week, Sinofsky talked about how the new version of Windows is designed to build on top of Vista's architectural changes without adding things like new driver models that can increase compatibility challenges. Below is the edited, but still rather lengthy transcript, of our conversation.
Q: In contrast to the pre-release publicity for earlier versions of Windows, we haven't heard a lot about Windows 7.
Sinofsky: We're always super anxious as engineers to talk about the work that we're doing. But on the other hand we really take seriously our responsibility of being part of the overall Windows and PC ecosystem. We want to make sure that when we do share information, that the information we share is accurate and reliable, and that we have in place the mechanisms for feedback such that the feedback is really taken seriously with respect to our plans. The reactions that we've had to some of the lessons learned in Windows Vista are really playing into our strategy of getting together a great plan for Windows 7, and working with all the partners in the ecosystem in a very deliberate way, such that the end result is a very positive experience for all of us.
How do you balance that with trying to make sure that people see a future in Windows worth investing in?
Sinofsky: Well, that's a great question. I think that when you say people, there are many audiences. What we're trying to do is be deliberate with each of the many audiences that we have to deal with, and give them the information such that they do see and share the optimism that we share for the future of Windows and the future of PCs. So, it's really an audience-specific type of question.
A lot of our readers are pretty passionate about computers, and we haven't heard as many reasons from Microsoft as we hear from your competitors about why people should be excited about the direction the
platform is going in.
Sinofsky: I would talk about an example of the way that we see this playing out, which is the way that we've talked about Internet Explorer 8. With Internet Explorer, of course, we have a lot of enthusiasts or very activist people who really want to learn about the future of our browsers, and we put together a plan that had some really significant investments, and we started to talk about them when we felt like we could demonstrate that those investments were going to really pay off, and that they were going to be actionable. So, you saw us do the planning, come through with working with all the partners that we work with and the influentials in a very deliberate, very one-on-one kind of way, and then we started talking about it broadly. We were ready with a beta, and we were ready for people to really act on the work we had done, and provided us the feedback that we're actually ready to absorb and ready to put into action.
So, for the enthusiasts, who are really excited about Windows, well, first, I share their enthusiasm. And second, we're really going to focus on making sure that when we talk about the product, that they're getting information that is really what we're doing for the product.
Do you think that makes it hard for Microsoft and its PC partners in the interim, as they're trying to sell consumers on Windows at a time where we're hearing a lot of ads speaking negatively about Vista, particularly from Apple? Is that a concern to you?
Sinofsky: There are a number of elements of the question, and certainly what I would say is when it comes to our partners, the people who make PCs, the people who make hardware, the people who build software, of course, our work with them is constant and ongoing. So, they're not surprised at all in the dimension of the things that we're doing; we're just working with them in a way that's specific to our different audiences. A great example of this is our enterprise customers, who do have multiyear plans. So with them, our salesforce is equipped to have this dialogue, to really talk about the future and the road map of our products. We think that for each of the audiences we have the kind of information that's required for them to act on it.
My question is, in the absence of information from Microsoft about where it's going, it seems like you have your competitor, in this case Apple, on the consumer front really defining Microsoft in the absence of Microsoft defining Windows out there in the marketplace.
Sinofsky: In a way that's a different question. That's sort of a question about how are we talking about our current products in the marketplace. I think that Apple has a very visible campaign, and we work with partners, and have a very different approach to how we're communicating our product. In a way, what I would say is Apple isn't really talking about where they're going, and that was the root of your question.
When you think about Windows, what does a good release schedule look like? We've certainly heard loudly from Steve Ballmer that he doesn't want to see a five-year time frame like there was between Windows XP and Vista. How often do people want a new Windows release, and what types of things should change from release to release?
Sinofsky: The way that I think of planning a release of Windows is--and Windows 7 will be no exception--we look at it as it's a major undertaking, and we're going to produce a major release of the product. Then what we do is we work on the plans, we get feedback from different partners at different times in the plans, and really the disclosure is when we start to talk about the information that's actionable and exciting about the product. The timing of it depends a lot on what we wanted to achieve, and you've certainly heard us, and we've been very clear, and will continue to say that the next release of Windows, Windows 7, is about three years after the general availability of Windows Vista, and we're committed to that, and we've signed up publicly to do that.
So, when Bill Gates
Sinofsky: What I think I want to say is what I just said, which is we said we'd be out there with a release of Windows 7 three years after the general availability of Windows Vista. We're excited; the investments that we have are really about producing a major and significant release at that time.
When you think about Windows, as the ecosystem and the installed base has grown so huge, it seems like the testing matrix and the list of possible interactions is so large that it's become very hard to change Windows. Do you think you can keep changing the operating system the way that you generally have, or does Windows reach a point where you want to basically take what Windows is today and run it in some sort of compatibility layer, and so you can really start fresh? I know that Apple a couple times in its history found itself wanting to do that.
Sinofsky: I look at it in a little bit of a different light. All of those IHVs (independent hardware vendors) and ISVs (independent software vendors)...I look at them as the key asset to the Windows and PC ecosystem. So, I don't at all look at them like a compatibility burden or challenge, to use the words that you used, but I look at it as well, that's the big asset that customers look to when they buy into a Windows PC. They say, hey, if I bought this printer five years ago, I want to keep using it, and I want to keep using it as part of my PC network. If I have this other piece of hardware, I want to keep using it. We do have to get better at the work that we've done, and, in fact, sometimes we make very, very substantial changes that are really multiyear bets.
A great example of that in Windows Vista is the work that we did on graphics. We did do exactly what you said would be very hard, which is we re-plumbed the graphics infrastructure for Windows. That has a huge number of benefits for the ecosystem at large. It means the drivers can be made more robust, they don't have to run in kernel mode and things like that. But we also didn't execute on that as flawlessly as I think we all would have liked collectively as the ecosystem. The team worked super hard with the partners in graphics to really do a great job, but the schedule challenges that we had, and the information disclosure weren't consistent with the realities of the project, which made it all a much trickier end point when we got to the general availability in January.
So, were the problems with Vista support and Vista enthusiasm--it sounds like you're saying they were mostly issues of disclosure as opposed to execution on the product. Is that right?
Sinofsky: I don't really want to dwell too much on the views of the past, and sort of just tell you again the lessons that we learned in working with partners. The team feels this tremendous responsibility to working with IHVs and ISVs and OEMs (original equipment manufacturers), because they're running businesses, they have their own business challenges, their own business goals, their own aspirations, and when we speak about what we might do, they will take it seriously. So, we appreciate that and we respect that, and it's a great benefit. But if we're not accurate or the information we provide causes them to do one thing, and then we change our mind, that doesn't bring the ecosystem forward. A big set of challenges that we learned...is making sure that the information we provide legitimately reflects the promises that we're making to ourselves and to the team as a product.
I know you said you don't really want to look back, so maybe looking forward a little bit...We haven't heard a lot about Windows 7, but we've heard about a couple of things discussed. The real areas I've heard a lot about are this idea of a new
kernel, a minimum Windows kernel that came up in a speech, and then some stuff around new user interfaces. Can you tell us a
little bit more about where those things fit in with how you guys are thinking about Windows 7?
Sinofsky: We're very clear that drivers and software that work on Windows Vista are going to work really well on Windows 7; in fact, they'll work the same. We're going to not introduce additional compatibilities, particularly in the driver model. Windows Vista was about improving those things. We are going to build on the success and the strength of the Windows Server 2008 kernel, and that has all of this work that you've been talking about. The key there is that the kernel in Windows Server 08 is an evolution of the kernel in Windows Vista, and then Windows 7 will be a further evolution of that kernel as well.
So, memory management, networking, process management, all of the security hardening, all of those things will carry forth, and maintain the compatibility with applications that people expect. Finally, we are going to make sure that the release is available both in 32 bit and 64 bit, which is an additional help for maintaining compatibility, particularly with device drivers. As the 64-bit ecosystem catches up, we expect more and more people, particularly enthusiasts, to be running 64 bit. For many people that's a great scenario today. I know I run 64 bit on most of my machines, including my primary laptop.
What was this idea then that got talked about in terms of a kind of minimum kernel?
Sinofsky: Well, why don't we stick at a higher level today, because I think that I don't want to really dive into the implementation details today.
Where do you see the biggest opportunities for the OS to matter in the coming years? One area might be new user interfaces, but people talk a lot about the browser making the OS less important. I know that's generally not a view that Microsoft holds, and usually not at all a view the Windows unit holds. So what are the ways that the OS can continue to matter?
Sinofsky: There will be a lot of features in Windows 7. It's a major release. I talked about the kernel and driver compatibility and (application) compatibility, but there is a lot more for us to talk about. We'll certainly be in touch.
Is one of the goals with Windows 7 that there will be more things right out of the box to get people interested in this release?
Sinofsky: Again I don't want to talk about any more specifics today, because we're focused today on how we're going to communicate things. But really again to really make sure I'm clear, we're working on a major release, and I think that each customer segment will have its own way of understanding what it means for them to be a significant release. And some of the things that we're going to do are going to make the release more applicable to a broader set of people, but it also might mean, oh, well, if you're not re-architecting the whole thing, then maybe it's not a major release. But we're actually going to bring forward the compatibility, and we're going to make sure that there's a lot of value for everybody who's a customer of Windows 7.
Are there any sorts of things that are happening in the overall PC world that are influencing how you guys are designing the operating system? I would imagine one of the things that certainly would influence it would be the sort of extensions to Vista, the Windows Live services. How important are online service extensions to the operating system going forward?
Sinofsky: The great thing about the Windows and PC ecosystem is that there's no shortage of activity in every dimension. Right now, today, we're really focused on just making sure everybody understands how we're going to talk about all of the things that we're going to do in this next release of Windows, but what we really want to do is kind of stay focused on that, and let us do a good job making sure that the things we are working on are really going to be great for each of the different kinds of customers of the product.
I think somebody that reads this conversation we're having is probably going to walk away again with an impression that I know you don't always like: that the Windows team is really being closed and isolated. I know one of the points you want to make is that you guys are talking to your partners. Is there more you would say about why is it important to be so selective about what gets shared ahead of time?
Sinofsky: So, everybody wants to know sooner than later what we're doing. And what we're always trying to balance is, well, if we make mistakes, then that has repercussions in the ecosystem that we don't really want to have, and we really want to be a responsible team as part of the overall ecosystem. If folks need an example, the best example that I could offer in terms of really trying to be respectful of the needs of the marketplace is really what we've been doing with Internet Explorer. I feel like we really worked on a great plan. The people who helped us to design how we were going to be compatible, how we were going to be compliant, the standard support that we did, were all part of the development process early on, all the outside parties.
Then we turned around and said, "OK, now we're ready to go to developers." We had a conference at Mix, and we talked about the development opportunities in Internet Explorer, because they were actionable. We gave people the code, we had published the specifications, we were ready to go not just for them to go do the work but for them to give us the feedback, and we were in a position to really act on it. That's really what we're trying to do with the next release of Windows as well.
I can understand why it's too early to talk about specific features in 7, but I'm a little surprised that it's too early to talk about some of your philosophies about where Windows has an opportunity to grow, and how things are changing, and some of those things, some of the factors that are influencing your work.
Sinofsky: Well, again I think what I would say is that we're talking about different types of customers, and different types of customers have different needs for information and are able to absorb it in a way that I think is mutually responsible.
If I'm understanding correctly, the things that you guys are ready to say about Windows 7 is it will be in 32- and 64-bit flavors, and the idea is that 64 bit will grow over time, although it's still kind of an enthusiast thing.
Sinofsky: Well, I didn't say it's still kind of an enthusiast thing; I did say I expect for sure our enthusiasts will be running it. It's actually professional graphics people who use it, industrial design uses it. There are a lot of segments that are very active in using it.
Do you think that 64 bit has come along slower than people would have thought?
Sinofsky: A lot of drivers haven't been written yet, and we expect that they're available now with new hardware, and we expect that that library will be built up over time.
Then the other thing that seemed like you were saying about 7 is that it will really focus on the underpinnings that are in Vista and Windows Server 2008, and so people should expect new features but not necessarily a lot of under-the-hood changes that require significant testing and so forth. Is that correct?
Sinofsky: I didn't actually say that. What I did say (is) that Windows Vista established a very solid foundation, a multiyear foundation, particularly on subsystems like graphics and audio and storage and things like that, and Windows 7--and then Windows Server 2008 built on that foundation, and Windows 7 will continue to build on that foundation as well.
You mentioned Windows Server 2008 being kind of the core on top of which you've built. Does that mean it gets some of the benefits of the modular architecture that the Server 08 release had, where you had this notion of a Windows Core configuration of Windows Server, and then you can sort of add pieces on top of that when they're needed?
Sinofsky: I think we've talked enough about the direction that we're heading with the specifics of the product today, since we really did want to focus a little bit more on just talking about how we're communicating with partners and customers and the ecosystem at large.
Do you think there's a risk that the more tight-lipped nature publicly will alienate some enthusiasts and folks who really want to know early on where you guys are going?
Sinofsky: There are many different models for disclosure that different companies work in, and I talked about the one that we're basing on the lessons that we learned from Windows Vista. But, of course, you could look at any of the other vendors in the marketplace, and see how they deal with disclosure, and come up with different models, and speculate about the pros and cons that they really see. I think that we're just focused--the No. 1 goal we're focused on is really the responsibility that we feel, and the respect that we have for all of our customers and partners, and making sure that what we share with them is really accurate and actionable, and that we are focused, like I keep saying, promise and deliver.
Let me just end with this. Look, we're working--the team is working super, super hard on this release of Windows, and you have to imagine we'd really be excited to start showing it to people. We want to show it, and we want people to get their hands on it, but we want to do that under the umbrella of being responsible members of the ecosystem, and being respectful of people's time and energy and the work that they'll put in to making Windows 7 great from the work that they can do.
So, why don't we say we're on target for the three years after general availability (of Vista), we're very excited about the release that we have, and we're very focused on promising and delivering.
Click here for full coverage of the D: All Things Digital conference.