Allchin: Windows Rights and Wrongs - Page 2

Linux Competition">

What are your fondest, and least fond memories, of your time spent at Microsoft and why?

There were so many good times here, the craziness, which unfortunately I dont think we do as much anymore. Some of them I really cant repeat.

We have had fun joking with competitors and internally, whether it be visiting the competitors—with the Windows 2000 launch we did some funny things with competitors and partners. [But] there is not as much fun that happens anymore.

I mean, good barbs from [Sun Microsystems Inc. CEO] Scot McNealy always helped energize us here and we would think about remarks we could deliver back to them. Those were some of the most fun times, doing things that grown ups probably should not do.

If you work on a project for a very long time you have got to have an outlet where you can have fun with others. I think it was the Daytona release [Windows NT 3.5] that Dave Cutler [who lead Windows NT development] brought his race cars onto the campus and we drove them onto the lawn and everyone got pictures of them. It was just a good time.

Recently I went and bought a big, heavy metal pig with wings and I brought it in to performance team meeting and said okay, it is now time to get this thing to fly. You know, just things like that. There is also a lot of camaraderie that happens if you go through hard times together that you can never replace, ever.

From your perspective, did the success of Linux and the growth in free and open-source software affect the way Microsoft developed Windows, the applications that run on it and the developer tools for it?

As you phrased the question, I would have to say no. We have learned a tremendous amount from them in terms of community. In terms of the development tools, there are some processes we have probably learned, but we can learn as much from frankly, as the Linux community in how efficient they are. But most of our learning has been internally.

Microsoft did not cut its teeth, so to speak, with Windows in the enterprise. How difficult was the move into that space, and how well do you think you met those challenges?

This is a glass-half-full, half-empty story. I think our reliability and resilience is excellent. I think our cost compared to others is excellent, but, if I look at the half empty, I can say, Wow, there is more we can do in terms of application compatibility, in terms of deployment and the like. So I think we have made progress across the board.

If you go back to the early 1990s, people would laugh about Microsoft having a server. It was viewed as a huge joke.

/zimages/2/28571.gifClick here to read about the latest progress on Windows Server R2 RC1.

Once we started to make progress with the product, they then said we didnt have any field infrastructure to support it, and then they said they needed consultants on-site, and we wouldnt do that. It was one thing after the other. So it has been quite a transformation.

In reading speeches you gave from about 1995 to around 2000, there is one noticeable and recurring theme: DLL hell, blue screens, application incompatibility and reboots. That was an extremely frustrating time for users. What were you and the development team doing to address this?

That is a very difficult problem, as DLLs offer advantages. The DLL problem was addressed by the work we did on side-by-side [implementations]. There is a set of technologies in the products today that let applications create side-by-side implementations, and the operating system can help support that. The way the Common Language Runtime and versioning is done now is pretty much side by side, so that you dont end up with that versioning problem.

In terms of the reboots, I consider that to be something we have analyzed six ways to Sunday, and there is a team working on this constantly.

There is an ongoing theme in many of your speeches over the past 15 years about having to drive down costs. When do you reach TCO nirvana, if there is such a thing?

Software is not the cost, its the operation. As with any of this stuff, you never reach perfection. So we have done a lot on reliability and people see that today. We are doing a lot on security and people will see that. I dont claim success there yet, but we are doing something there.

I claim that we will do something about deployment, although we have improved things quite a bit, the work that we are doing in Windows Vista in terms of the image management, is dramatic and original equipment manufacturers as well as corporations will appreciate the cost savings of the image management. But there is a pretty significant leap that can be done in the future on that.

But there are two parts to this equation: one is about cost and making the system more efficient, and the other is about adding value. Above-the-line you see things like what were trying to do with the new graphics system in Windows Vista, where we are putting a technology that others can take advantage of.

There are many things that will happen in the future on this front, and this is to some degree about us being creative and inventive and taking advantage of innovations that happen. It is not like we invented RSS, but we are creating a platform so others can take advantage of it.

Also, one day we will have speech and as we get a platform set of features in, then applications can take advantage of this and we are making a difference and there is value on top that we can add.

No look at the development of Windows would be complete without a look at the Trustworthy Computing initiative. How has it affected the way Windows developers program and the way Microsoft delivers more secure products?

Before we introduced this initiative, I guess I was naive about the level of issues we would see in the future. I dont think Im being naive now, as it is front and center of everything we do, and just like those reboots were a problem, feeling safe on your computer is my current hot button, and certainly the work we did in Windows 64-bit and SP2 [Windows XP Service Pack 2] are examples of that. Vista will take a huge step forward.

People may say that Trustworthy Computing is just a big label, but we swung the group around to do something about it, and we still have a ways to go.

The U.S. governments antitrust case against Microsoft played out in the courts all the way from 1997 to late 2001. Can you now, in retrospect, tell me what the biggest impact of the trial was and what you learned from it?

Well, it was very hard on me personally and on the development organization [and] the company. The integrity question is a very important one for me as well as for the company, and people here knew that they were working for a great company and so there was certainly a concern about how things were being portrayed. There were many [things we learned]: ... Transparency and leadership were good things for the company to learn, in my opinion. They were painful, and I dont think it was necessary, but those things make us a better company today, and my push for doing CTPs [community technology previews] is not related to that per se, but there is probably a [spillover] effect.

It is hard to have controversy if everyone knows what you are doing, right or wrong, so we are doing a lot more talking about everything we are doing.

Why have you have stayed at Microsoft for so long, especially in this, an industry where people are always moving on looking for excitement and new challenges?

I believe, because I believe the impact Windows can have is so huge. If you mess it up the impact on society is huge and if you dont mess it up, the impact is also huge. So Id rather stay and do a great thing.

I look at what Windows Starter Edition has already done in countries like India on the education front and I like being part of that and feeling that I am doing something good for society. I am not someone who will cure cancer, but maybe I can provide equipment to someone who is.

There is a database being used right now using some of the stuff we put together for the Red Cross after Hurricane Katrina. Thats why I stuck around. Yeah, there were lots of other things that I could have done, but we have impacted society in a very good way.

When you finally hang up your Microsoft hat around the end of 2006, what would you most like to be remembered for?

I havent thought about that, you are hitting me a little early. A believer in quality, a driver of quality, and a member of the team.

Editors Note: This story was updated to include further comments from the interview cut from the print version for space considerations.

/zimages/2/28571.gifCheck out eWEEK.coms for Microsoft and Windows news, views and analysis.