When Microsoft corporate vice president Steven Sinofsky appeared before a hostile group of anti-Microsoft journalists at the D9 conference to unveil the Windows 8 Start screen, I was transfixed. But I was transfixed virtually since D9 didn't broadcast video of the event live and I instead had to rely on what turned out to be a miserable live blog of the event that I used as the basis for my article Steven Sinofsky Live: First Windows 8 Preview. I've long decried the biased nature of the folks from All Things D, but D9 was a new low, and now that the full, unedited video of the Sinofsky interview is available (a dubiously edited version was previously offered, hiding many of the most important bits), we can finally find out what really happened.
By the way, I'm not alone in being unimpressed by how All Things D handled Mr. Sinofsky. My Windows Secrets co-author Rafael Rivera took the time to transcribe the edited video and then compared what Sinofsky said to the horrible live blog account, which left out important information while providing a bunch of spurious Microsoft digs that had nothing to do with the actual event. Please make sure you read his post, D9's Odd Windows 8 Video, Unofficial Transcript, before proceeding.
My goal here is simple: To discover what Sinofsky really said at D9, especially about Windows 8. So this is a much more accurate transcription of his talk, complete with almost every unnecessary Walter Mossberg interruption (and there many). I skip out on some of the opening bits, and then get more detailed and accurate as the talking continues. Mossberg's comments and interruptions are in brackets [like these] and his comments are in italics like these.
When asked about his role at Microsoft, Sinofsky mentions he is in charge of Windows, Windows Live, and Internet Explorer. He previously worked on Office for about 12 years. "I just like to think of it as cool software," he said when Mossberg mentions that these two products represent much of Microsoft's market power.
When asked about the size of the Windows team, Sinofsky notes that there are about 1400-1500 software design engineers on Windows, Windows Live, and IE, and then testers, program managers, about 6000 people overall in the organization ... including all the marketing and outbounding work.
OK, let's dive into the meat of it.
[How do you feel about not being in the gang of four?]
I'm watchin' it, I feel like Microsoft is auditioning on The Voice, we're just waiting to get picked. Whatever.
[15 years ago, we might have said there was a gang of one. ... blah blah blah.]
Well, nothing called the gang of four ends well, so I didn't really want to be part of that one. There's just so much going on and the thing that struck me the most was, it's an interesting group of companies, lots of creative work, but the one thing that gets left out of that kind of discussion is the way that 90 or 95 percent of the world gets on the Internet in the first place is through Windows.
[It's not in mobile in a lot of countries?]
Well, it is in mobile if you look at metrics ... there's a lot of usage of Windows. Leaving that out is an odd way to gerrymander the gang.
[Microsoft have missed a couple of things. iPhone...]
Well, we did miss the iPhone, we didn't do that one. We have a new entry, we're trying hard.
[Continues... you're years late on that, still not seeing tablets .. iPad 16-17 months ago ... you did tablets before, didn't succeed .. what's going on? Why? Microsoft has smart people blah blah blah on and on he talks .. Are you big and bureaucratic?]
There are always things we're doing well, and thing we're not doing well. You picked two of the things that we didn't do particularly well on. And it doesn't mean we're out of the game, it's a long game, and you don't want to write people off too early. It's only been 16 or 17 months ... on the tablet. [Mossberg interrupts, a little later on the phone] We were later on the phone, but it also had more success at the time at a relative scale in the world of phones. It's just more opportunity for us to do a better job. There's not some magical answer to that. We didn't do the job and now we're working.
[Don't you think there's a systemic issue of (not) being nimble at Microsoft?]
I don't really think so ... when something comes out, then everyone wants the immediate response. But things are always happening in parallel. There are lots of smart people at other companies and it turns out we're all sort of ... [Walt interrupts] Exactly, I'm saying ... We're all looking at the same problem space and pulling in many of the same components to build solutions, and so then we're all piecing them together not just different ways that make us all unique, but some of the ways work, some of them don't work, but also on different rhythms and different time scales. The world of browsers looks a lot like that. People are doing some things and other people are busy working and someone releases and then some time goes by and the other person releases. But it has been only 16 months on the iPad. Certainly phones have been a little bit longer and we aren't there. So we're just going to keep working and keep trying. And I think here today we'll talk about some of the things we're doing certainly on the touch user interface world, the touch world.
[Talk about that. Microsoft has two well-known OSes. He makes a Windows SKUs crack. Do you want to defend that?]
Can I have a chocolate?
[You have Windows, still by far dominant OS. You have Windows Phone 7, which is newer and your entry in the super smart phone game. How do you get to the tablet? Down from Windows? Up from Windows Phone?]
What we are doing, and what we're going to talk about today, is how we're going to build on the flexibility of Windows, the Windows operating system that's in PCs today, and extend it to a broader range of usage interaction and a broader range of platforms, if you will, or form factors. The thing we really looked at ... we started this release of Windows that we're going to talk about today, we took a step back and we wanted to ... you know, what are the assets we have and what are the best ways to move forward, and what are the customer challenges that people have, and ... one of the things, for me, is striking is that over the 25 years of Windows, has been the flexibility of not just the Windows product but in a sense the Windows code base, the Windows approach.
When you think about what it was like when (Windows) started, the world was CGA monitors and giant machines with, maybe, kilobytes of resources. And over the years, Windows has gotten richer, and it's used more of the Moore's Law of hardware, but the scenarios it's attacked and the ways in which it's gone after those scenarios has also changed dramatically.
[Give me an example]
We've gone from ... even the transition from desktop to laptop [explains]. Another one is how we went and took the same Windows operating system kernel and made it go from a single processor, 100 MHz Pentium up to these 192 processor, gigantic, datacenter servers that we see today. And it is just the same operating system kernel. And when you think about user interface, we built Media Center, we built the Pen and Ink interface ...
[Interrupting, To name two that ... basically failed]
Well, it depends on what you want to measure as success or failure. Certainly, they're all used by lots and lots of people.
[They're not mainstream successes, big hits. Interrupts repeatedly as Sinofsky tries to address the topic]
Once you get into a feature of Windows, then any one thing is going to get used by some percentage of the base, not by (everyone). There are only a couple of things that every single person does. But it's that sum total that's really the flexibility of Windows in the first place.
So when we looked at what was going on with tablets, or with touch user interface, we sort of just asked ourselves, what else can Windows do? This is not some giant leap. We certainly worked on touch, you know we were here a couple [three] years ago showing touch in Windows [7] ...
[Walt interrupts again. It has it in a little different way, using your finger instead of a mouse]
Exactly...
[Interrupts again]
That was our design point, could we attach a touch digitizer to a monitor, in a laptop or something else, and then have that work as well as a mouse? Turns out, that probably wasn't the best choice to the design point. At the time, there wasn't a big touch slate anywhere ... so we thought it would open up some new scenarios. We were particularly interested in large screens, in a classroom, or a Point Of Sale ... and it works for that super well. It doesn't work as a replacement for the mouse and we never really thought of it as a replacement in Windows Explorer or even in browsing itself. We did a bunch of work, but clearly the world changed underneath it. And whenever the world changes, we take a step back and say, hey, what else can Windows do to change to meet those scenarios?
When Windows started, there were almost no computers connected to a network at all. The difference between Windows 3.0 and 3.1 was sort of the whole introduction of networking into the PC.
[But that was local area networking]
Well, and then Windows 95 introduced TCP/IP and broader Internet networking as well. And the Windows NT kernel introduced server networking, and all the work that goes with that. So this flexibility sort of forms the hallmark of what we look at when we start to build a release of the product.
[Walter interrupts again, pontificating at great length: I've always been told that Windows is this gigantic, enormous ... thing, with all this legacy stuff we've had to keep in there. Some of your competitors have not, but you have all these enterprise customers, software that's very old but still has to run, so it's this big, bulky thing. Some of us have had that sense on our own computers, that the longer we've had Windows, it seems to slow down and drag... you just have this sense it's a big, sluggish thing. When you flip on your iPad, or your Android tablet, it feels light, it feels new, it feels extremely responsive. You really think ... you have this problem, and it's an ironic one. You were doing tablets for a long time and they didn't catch on. The guys in Cupertino did a tablet and somebody said, "it's rocking," great. You want to be there. You want to be in that arena. Why would you turn to this big, heavy Windows thing instead of doing another OS or a shrunken down version of Windows or a scaled up version of your phone OS?]
Everyone said the same thing about servers, so why not write a whole new one, because it doesn't have enough stuff...
[Walt interrupts yet again, that didn't happen, not everyone cares about servers, blah blah blah]
It turns out servers are even more important now
[Interrupts yet again, Really important, yeah I know]
In a sense, I think it's sort of a fair question, but an unfair question, and I signed up for either.
The thing that's most fascinating about the evolution of Windows over the past few decades is that, yeah, it grew up with hardware. So the hardware companies are busy doubling every 18 months, and adding less fat, and we were doing that with the software part of it as well. And at some point, we reached a plateau, in terms of, we don't need more of some kinds of the resources ... and there are other kinds of resources that we needed more, and what happened was, we were doubling the system requirements for base level Windows every release, which turned out to be every three or four years or so.
[Vista was a very good example of that]
Well, every release was like that.
[Well, my recollection was Vista required a particular ... if you tried to put Vista on your existing computer, it didn't work very well. You really had to have a new computer.]
Fine, OK, so yes ...
[Interrupts again, You weren't running Windows at the time]
Sure, that might have been worse, or better, whatever. What happened with the transition from Windows Vista to Windows 7 was that we looked really hard at this and we just said, wow, we can do a bunch of work so that we don't need to change the system requirements for the release. And so Windows 7 was the first release ... actually, they went down if you look at the three main ones, which were graphics card, memory and CPU and then disk footprint ... For the first time, we actually went down on system requirements. But a very interesting thing happens along the way, which was .. What were the system requirements of a phone? Or a super smart phone? Or a personal computer in your pocket with a four inch screen or whatever? Turns out those (requirements) have been doubling (too), but every nine months. And so we've gotten to the point now where the phones actually have the system requirements for Windows in them. So it's not some giant ...
[Walt interrupts yet again, heaven forbid his point be disproved, They don't have the space for Windows]
Well, they do. When you look at the system requirements for Windows 7, or the ...
[Another interruption. Then why don't your phones just run Windows? Why is there this thing called Windows Phone?]
Well, we're doing a bunch of work now, one of the big pieces of work we've done is work on the chipsets that are used in most phones, which are a different instruction set called ARM.
[Walt interrupts to say "ARM" over Sinofsky, like they're having a race to get to answer. Sigh.]
That's something we've shown for Windows already, is that working. In fact, we're in Taiwan in about two hours or so, and we're going to do an update for all the work we're (doing) on ARM, and we'll show a bunch of new devices that are prototypes that actually look like tablets and look like new form factors running the next release of Windows on ARM.
So we looked at those system requirements, we looked at the ARM processor and what's available, and we looked at the Windows code base, and we said, well, they're in sync. We can do this work, and we can continue to drive the base system requirements. You know, most phones have 512 MB of RAM, all the slates have 1 GB, and the ones we're looking at in Taiwan look like they're going to move to 2 GB, and that is a base level Windows machine. When you look at the RAM usage, the number of processes, the disk footprint, those are all the same as what you see on the current crop of slates.
[So full (new) Windows, which will be called...?]
Ah, yeah, well. So we have this whole thing about what to call the next release of Windows. So we're just going to call it a codename because I love codenames. No, we're just calling it Windows 8 right now, for the purposes of this. We had this whole debate in Office for many years over what to call the release under development, and so we had these big meetings, what should we call it? And we had codenames. Should we call it Firestorm, you know, Death Ray?
[Your codenames are always better than your product names .. for some reason, this gets a laugh.]
What was it, Personal Photo Twitter Delivery Service Something?
[Stupid side-joke about Twitter and gang of four ensues, briefly.]
We started just numbering the releases, the same as, sort of, what we were thinking, so you know, Windows 8, with air quotes, and give us some time and we'll figure out the actual product name. Which we'll do, at some later date.
[This is not the strategy that's been followed by your two biggest competitors, Google and Apple, who have ... blah blah blah. Side conversation about whether Google has a "legacy OS" and Walt finds a reason to pontification some more. He always does. Eric (Schmidt of Google) said one's for touching (Android), and one's for typing (Chrome).]
And you bought that?
[Walt then incorrectly summarizes Apple's strategy, claiming they didn't move OS X to the iPhone. Why is (what Microsoft is doing) a better thing for consumers?]
It's better because of all the things Windows brings with it, not just the ...
[Walt interrupts yet again. This guy can't let anyone answer a question. Here's where I can be snarky and mean, and say viruses and craplets and all the things that people get on their Windows computers.]
Or, you know, printing ... and using solid state storage really well, and external hard drives, or ...
[Another interruption. The other guys use solid state storage pretty well]
... without weirdness and stuff like that.
What we set out to do with Windows 8 is really reimagine how to work with the PC, and really think from the chip all the way up to the user interface what it is that we could do different(ly). Not just to get touch, and get a slate or a tablet made, but to really think about the broad range of scenarios and really reimagine what we want to do with Windows. You could say we colored outside the lines in terms of how we built this release. [Walt tries to interrupt again, but Sinofsky, finally, ignores him.] So we're excited to talk about it because we think it is a different approach ... We chose a different avenue. Maybe we have real customer and engineering scenario things that we thought about. We have an approach that is different, but also builds on the value of, you know, 400 million PCs that will probably be sold in the year we release this product. That's a big number, and what will happen is, all of a sudden, all of those PCs accrue all of the benefits of the work we're doing...
[Mossberg interrupts So both tablets and PCs, laptops, desktops, whatever ...]
Right.
[Interrupts again]
So what we've done is, we've looked from the ground up, how to rethink about how you interact with Windows, the kinds of programs you can run, how you get those programs, all that work, and really bring it in ... a word we used a lot in developing it was, modern .... how to think about it in a different way that really solves a bunch of the .... Things that people see or say they say are solved in an iPad. And I think we can do that and then bring with it all of these benefits you have...
[Walt interrupts again. So every program that runs in desktop Windows will run on ... waves hand]
Yeah, it's Windows. Windows is there and everything that runs on Windows 7, every device that you can plug into a Windows machine, everything runs.
[OK, Walter says disbelievingly.]
From here, the demo ensues. But that will have to wait for Part 2 of this transcript, coming soon.