|Craig Mundie: set to take over from Bill Gates in July 2008|
Bill Gates has left the building â almost. In July 2008 the self-made billionaire will make his final exit from Microsoft. Gates will focus on his philanthropic activities and his role will be taken over and shared by two Microsoft veterans: Ray Ozzie and Craig Mundie. Ozzie will be Microsoft’s Chief Software Architect looking after technical strategy and product architecture but Mundie has the role of setting Microsoft’s strategy across the board as well as heading up all of Microsoft’s research and development.
In this exclusive interview with APC Mundie says the notion of all software delivered entirely through the web browser is now widely recognised as being ‘popular mythology’. He also stakes the claim that Google’s existence and success was contingent on Microsoft creating Windows. He talks about what’s coming down the pipeline for future versions of Windows and his belief that Windows can get still more market share than it has today. He also discusses the issues around the recent controversy over the Office Open XML file format.
Craig Mundie spoke to APC’s Online Editor Dan Warne.
APC: Craig thank you for making the time for APC readers. It’s a rare opportunity to talk to the person who’s taking over Bill Gates’ role! I understand you are taking on half of Bill’s role with Ray Ozzie taking on the other part. What does it feel like to be taking over from the most powerful person in technology? And what passions do you feel you are bringing to the role?
Craig Mundie: Well I have been Bill’s closest partner on many of the long term technology strategy and research projects for almost ten years now. So I have quite a bit of personal familiarity and share many of the same passions that Bill does. I’m passionate about the importance of sustaining a long term investment in fundamental computer science research.
As we partition Bill’s role between Ray Ozzie and me the part that I particularly picked up is the management of our global research operations and where we maintain the focus on investigations of these longer term concepts.
APC: So what are your observations of the tech industry after having worked in it for so long?
Craig Mundie: Well I think I have learned quite a few things but I guess one is the importance of patience. I came to Microsoft 15 years ago to start ‘non-PC’ computing. At the time Bill could anticipate that the microprocessor would get more powerful and shrink physically and we would end up putting it in a lot more things than just PCs.
Of course then the world was strictly focused on the desktop computer and traditional IT kind of environment. So today 15 years later what we started in 1993 is only now coming to substantial market fruition.
One of the things that I’ve learned is it is easier for us to think about the technology and how it is likely to evolve than it is to get the world to quickly move to take it up at scale.
I think one of my big learnings has been that it is a big challenge to change society’s infrastructure. That has been really what Microsoft has been a big part of doing in the area of IT globally. When you look at each new area whether it be smart phones or interactive television or essentially another generation of game consoles we have been able to understand what it takes to build these things even though we always face a substantial challenge to get the critical mass.
APC: Vista has been criticised for having too few improvements over XP. However having followed the betas closely and interviewed some of the Vista developers what is evident to me is that there is an enormous amount of functionality waiting to be tapped like the Windows Communication Framework and Windows Presentation Foundation. How long do you think it will be before the developer community hooks into Vista’s Ohidden features’?
Craig Mundie: The way I think of platforms and their evolution broadly is that they always go in cycles and the cycle has two components or two waves to it. The first wave I call the diffusion cycle and the second the exploitation cycle.
Developer communities really only function in the large in the exploitation phase and the platform has to largely be established at scale before people move to really capitalise on its intrinsic capabilities.
So the diffusion cycle is always driven by a small number of similar applications or things that have broad appeal. That tends to get the platform out there and once it’s out there you can then educate people about these other benefits and they essentially move gradually from their platform model to the new platform model.
Microsoft has done that successively over and over again even in the evolution from DOS to Windows then to Windows NT-based technologies and now on to Vista.
The same thing is happening between generations of phones game machines and IPTV technology.
So Vista is in its diffusion cycle and until there is enough of it out there you won’t really see the developer community come across.
So far we have delivered about 60 million copies. That would represent about six per cent of the global Windows install base. So it has probably got to get up another few percentage points before you will start to see a big migration of the developer community. But I think that the tools are there to support that. More tools are coming online that I think are also important like the Silverlight technologies and of course the .Net framework which has been broadly adopted now.
So those things come together and then you get this big shift. I don’t think it will take many many years but I don’t think it will happen in the next few weeks either.
APC: So Microsoft’s big money-makers are still Windows and Office right? Is there going to be an inflection point where Microsoft stops relying on desktop software revenues and focuses on online services?
Craig Mundie: No I don’t think so because our view right now is that the future is really about these composite experiences of locally running software in the client devices working in conjunction with services that exist in the internet cloud.
Some of those services are infrastructural and some of them are experience specific. We have been working for the last few years to add the service component to both the platforms in almost every single Microsoft software product.
I think there was a period of popular mythology where everybody was saying the whole world is going to migrate to ‘software as a service’ over the internet. We absolutely don’t think that is true and I think it is increasingly recognised as not likely to be true.
Rather what will happen is that you’ll have a seamless integration of locally running software in increasingly powerful client devices (not just desktops) and a set of services that work in conjunction with that.
A lot of what we are doing with the Live platform not only allows us to provide the service component for our parts but also gives the abilities for the developer community to perfect their composite applications and get them deployed at scale.
APC: So do you feel that is a major competitive advantage over Google for example? That you have the desktop software expertise along with online services?
Craig Mundie: Well yes and no. I mean the fact is: Google’s existence and success required Microsoft to have been successful previously to create the platform that allowed them to go on and connect people to their search servers.
Now Microsoft’s business is not to control the platform per se but in fact to allow it to be exploited by the world’s developers. The fact that we have it out there gives us a good business but in some ways it doesn’t give us an advantage over any of the other developers in terms of being able to utilise it.
For example as much as our Virtual Earth product uses a lot of local 3D rendering technology so does Google Earth. So I think there will be other ways in which we distinguish ourselves and where our knowledge of the platform and ability to continually evolve it will be a business advantage for us.
It is just the difference between being part of the infrastructure of the internet as well as competing directly in the service or client capability as well.
APC: I wanted to ask you about the problem of legacy software in IT especially because the new Office Open XML Standard was rejected today and some of the complaints were around legacy aspects of it like ‘Word 95 character spacing mode’ which Microsoft was unable to articulate as part of the standardE
Craig Mundie: We should talk about some of these issues…
APC: Sure but the question I was going to ask was is there some merit in Apple’s approach of simply cutting off legacy compatibility quite frequently?
Craig Mundie: You know it’s a luxury to be able to leave your past behind and if you have a relatively small installed base and you are essentially moving more into consumer-oriented things you can be a bit more of a cavalier about that.
What the world looks to Microsoft for is the stable infrastructural component and many core technologies that run other systems. If anything they want us to be very methodical and discerning about how we move forward and the rate at which we obsolete the things that we have left behind.
We have been under continuing pressure from our customer base to extend the life cycle of support for many of our products and have done so for as long as a decade on many occasions.
That just reflects both the depth and breadth of the use of the technology and the fact that people don’t want it to change out from under them very dramatically. I think they also recognise that there is a price to changing things in terms of the retraining of the consumer experience.
Vista and the new Office System represent some of the biggest changes we have made in some of the concepts around user interface. It is one of the challenges we have to get through; that initial psychological shock of trying to move people forward from that which they have become comfortable with.
Our business is much larger than Apple’s and embedded in places where the stability is viewed as an asset and so we just can’t really leave it behind as freely as others might.
APC: So getting back to the Office Open XML StandardE
Craig Mundie: Yeah so a couple of things. Firstly the standard wasn’t quite rejected. We are in the middle of a long process which won’t be completed until Spring of 2008 where the final decision regarding standardisation will be complete.
We were actually quite enthused with the number of countries even at this stage in the balloting process who voted to approve the standardisation at the current level. While there were a number of countries who either abstained or voted procedurally ‘no with comments.’
So we have now entered this ‘ballot resolution meeting phase’ where the standards body drives a process which we obey and are committed to to work through the comments that were raised by people as we move forward in this area.
There are a lot of people who have raised a great many issues which we don’t think have a lot of practical merit but serve the purpose of creating some anxiety during this process. Many of the comments that were submitted had common threads and were put together by people who oppose this activity.
So given all of that we were actually quite positive about the fact that we came within literally a couple of countries of having the thing ratified even at this point.
We are fairly confident that having looked at the comments that certainly most of them will be readily resolved and that a lot of the countries who voted ‘no pending comment’ are likely to switch to a yes. We literally only need to get a handful of those to convert in order to get ratification.
We view it as a relatively positive outcome. We certainly would have preferred that it had gotten ratified at this point but we are not at all deterred from thinking that people will ultimately endorse ISO standardisation.