News Stay informed about the latest enterprise technology news and product updates.

Microsoft's Taylor-made Linux strategist

NEW YORK -- Microsoft Corp.'s platform strategist Martin Taylor came to LinuxWorld Conference & Expo with a mission -- and a flak jacket. Taylor jokingly wore the bulletproof vest during the first few minutes of an hourlong public question-and-answer session at the company's booth on the show floor. Taylor faced the open-source enemy head-on and answered audience-submitted questions and winked at a heckler or two. Shortly thereafter, he spoke for 25 minutes to about Linux, which he said from a platform perspective is "the biggest competitive area that we think about." Taylor also talked about Microsoft's new "Get the Facts" ad campaign that tries to dispel the notion that Linux has a better total cost of ownership over Windows. He also answered questions on Linux's accelerating adoption rates, why Microsoft is pouring resources into combating Linux, the challenge Windows security raises for Microsoft in conversations about Linux, and a lot more.

You understand the perception that Microsoft is somehow pulling the strings behind SCO's lawsuit against IBM?
I wish we could be that smart, that Machiavellian in our approach to doing business. No, I really don't. We would never be and we won't be. That's not how we do business. I think SCO has an opportunity to look at their installed base and say 'Where are these guys going to go? To Linux, or Windows.' I'd love for things to happen, but SCO is really driving their own business model and God bless them. We're focusing on Windows Services for Unix. How tight is Microsoft with SCO?
We have a relationship through our Services For Unix product, and outside of that, that's the extent of our relationship. Do you see any merit to their claims?
I'm not the right guy to talk about that. I didn't go to law school.

FEEDBACK: Is Microsoft's posturing against Linux panic?
Send your feedback to the news team.

Do you ever see the day when Microsoft would buy a distributor like Red Hat the way Novell did with SuSE?
I don't know what it would gain us. I'll tell you why Novell did it, because they had an operating system that was losing market share and they need to provide something else in order to inject life into their installed base. They have a set of management tools that they want to integrate into something. I know why Novell did it. I'm not sure why [Microsoft] would do it. Why would we? Turning to licensing in terms of the city of Munich recently announcing it was moving off of Windows on the desktop to Linux, what strategy changes have to happen to prevent more Munichs from happening?
Munich had nothing to do with licensing, and had everything to do with politics. It had everything to do with making sure we were engaged with the right political parties the right way. It gets broader than that. It gets down to making sure we are a responsible citizen in all of these countries.

Many of these countries are not looking for Microsoft to sell or give them software. They're looking for Microsoft to really help them build a society of people who are technically literate. They want us to help them build an infrastructure that helps them participate in this IP-based economy, versus a more manufacturing, labor-based economy.

Munich wasn't a technical discussion. It wasn't a pricing discussion or a licensing discussion. It was about whether we were connected to the right people, the right parties at the right time. To make more money, or perhaps cannibalize Red Hat's business?
I don't see the day. Our bet is on the Windows platform. We know we deliver a hell of a lot of value on the Windows platform. We know that we are approaching 10 million servers in the marketplace. For us to buy a Linux distributor would mean that we no longer saw value in the Windows platform. That's the last thing we'd ever want to say about our company. With the pervasiveness of Windows, isn't it incumbent upon Microsoft to close those vulnerabilities quicker. The vulnerabilities are real, there are critical patches released monthly, all of those factors weigh against your argument.
Again, you have to separate the noise from the reality. One of the GNU servers that holds the Linux source code was hacked. It was not very public, but of course, that's like someone hacking into the Windows code base. It's an industry-wide problem, it's not just a Microsoft problem. I feel great about our progress.

Our vulnerabilities are actually decreasing, not increasing if you look at CERT data. Linux vulnerabilities are actually increasing. I respectfully disagree that we are getting worse. That said, we have a lot of work we need to do. I would say that in order of priorities, if security is not at the top, there's hardly anything else in front of it in terms of how we think about architecting better for security in terms of design and how we build things.

When it comes to deployment, we're making sure we're providing customers with prescriptive guidance on how to build secure solutions and also turning off by default things that don't need to be turned on to eliminate back doors into our code or technology.

Also, if you look at in terms of when a vulnerability is found and when a patch is available, and I think you'll be surprised in our ability to deliver that to customers versus the open-source community's ability to deliver patches to customers. That's primarily because we do the patch and the distribution, where the [Linux] community might do a patch, but Red Hat, for example, needs to do the distribution. So you have a much bigger lag time between the two. Is security a sticking point for Microsoft when it comes up against Linux. Do you hit a wall there?
We definitely don't hit a wall with security. It's the same discussion. The perception is that we're less secure. The reality is that we have more market share, so when we have an issue, it's broader. If you look at the vulnerability database, you'll actually see that the number of vulnerabilities is lower on our platform, both in absolute number, or take a particular time period and see how far down we've come. You have to come to a facts-based discussion on security, and not one just based on perception. I think that probably the area we need to continue working on is the high-performance computing area found in academics, automobile research, oil and gas, etc. Part of that is a licensing issue. If you need to deploy hundreds and hundreds of Microsoft servers versus hundreds and hundreds of Linux servers, it's a different licensing approach. The adoption rates have been so quick. It's moving off the edge and deeper into data centers.
Why is it moving? If you take a look at the Unix-to-Linux migrations, it's about the Unix equals Linux, Linux equals Unix connection. It's a skill-set issue; people know how to work on a character-based mode. They know how to write in Perl. We have Services for Unix that people can use, but most of your diehard Unix system administrators don't even want to touch a mouse or a GUI.

To the bottom-tier area, that's where it's more of a footprint, disposability issue where we offer a broad set of solutions and you pick what you need. With Linux, you can build a single-purpose server and do these things very easily. In those two environments, you don't see too many big, complex stack-integration Linux solutions. That's where it gets a lot harder for them and a lot easier for us to communicate our value. We have an ability to work with our partners and our solutions to integrate up and down the stack. Part of your job is to learn about and understand Linux. In your opinion, why is it succeeding?
First, define succeeding. Some of the research cited in 'Get the Facts' has been funded by Microsoft. Why use that research? Doesn't that taint your campaign?
I've told every single analyst firm, IDC, Gartner, Meta, if you guys did this on your own, I would not fund it. I want this data. Customers want to see this information. The fact remains is that they can't always fund it themselves. That's why [Microsoft] does it.

In terms of the credibility issue, I'm very transparent. I funded the research, but it's their process. The Giga-Forrester thing that's out there; it's their model, I did not create the model. I didn't architect the model. We're transparent on that. I encourage customers to go do their own. Take their model and replicate it. Hire Forrester or Giga to do it for you. It's so much less about the exact numbers, it's more about saying 'Don't believe what you think or might have heard.' Actually take a fact-based approach from analysts or do your own analysis to take an educated, informed decision. Aren't IT pros smart enough to do the research and get the answers to those questions themselves? Is it worth the resources Microsoft is putting into this?
It's 100% worth the resources. They are smart enough, but they don't always have the capabilities to do it. For instance, as an IT professional and I'm thinking about a simple solution for a small department, I'm probably not going to go into a detailed analysis about two platforms and all the time it takes to run the specs and run the [performance tests] to then have the comparative analysis to then make a decision.

I'm normally going to lean one way or another to deploy that solution, then make a decision based on what I think. If it works, stay the course. We have many customers that might be doing an internal TCO discussion. Others say they're going to take a look at what's out there from publicly accessible information to build their view. If the general perception is that Linux is winning against Unix in the server market, why all the positioning against Linux? Some might presume that Microsoft is panicking, and sometimes perception becomes reality.
I don't see it as positioning against Linux. Most of the migrations are against Unix, no question. There is also a little bit of penetration on single-purpose servers and appliances. Our focus is to make sure that we clear up the perceptions that exist.

You do some research and come back to me and tell me there's a perception in the marketplace that Linux has a better TCO. No, it's not the case. Let's do some research and make sure people understand that. In many ways what we're doing with the 'Get the Facts' campaign is making sure that on these areas where there is a disparity between reality and perception, we're trying to put those two things a little closer together. Are you hearing more business-related or technology-related questions about Linux?
Most of the questions are about interoperability and things like that. People might say Linux is a better product -- and that's not me speaking -- in terms of innovative solutions, new ways to do things, new ways to implement them.

We know we've got a great offering. More so on the business model, on the licensing. The loudest technical area is with Unix migrations. And even with that, it's not so much purely technical. Part of it's a perception issue and people giving Linux the credit that perhaps Unix earned in many ways.

So some of the challenges on the x86 architecture, you're going to see [migrations to] Linux or us. We've been at this for quite some time optimizing our platform for those migrations. So, I would say that's probably where most of the technical conversations might come up. Some research indicates that Linux is winning against Microsoft in some markets. Its acceleration and adoption is undeniable. What worries you about Linux?
I don't say what worries me about Linux. I think about the market share. When you see the growth numbers, you really want to understand why. Is it because we don't have the right technology that customers want? We don't have the right business model? It's no different than any other competitive discussion. It's the same way we think about Sun, Oracle, IBM, AOL…

It's about 'Do we have the right offerings for our customers?' If we do, are we communicating it the right way, or building the right things? We are thinking about it holistically. Linux is a slightly different phenomenon based on the business model. This is the first time we've 'competed' against this business model of low or no-cost acquisition.

Dig Deeper on Linux servers

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.