I first met Ian Murdock, now head of Sun Microsystems' operating-system platform strategy, back in 2002 at a conference. I was a bit in awe at the founder of Debian, only to find out that he's a very genuine, very likable person. I've kept in touch with Ian through the years, and on Monday was fortunate to catch up with him to find out how his move to Sun has gone for him, and what he's brewing.
Give me a quick update on life at Sun.
Things are going very well. It's a bit overwhelming at times. I've been a start-up guy for most of my career, so there's a certain amount of culture shock involved in working within a 33,000-strong company, and understanding how to navigate the machinery and history of the place.
I came to Sun with a range of well-defined ideas. My questions revolved around actually seeing them through to fruition. How was I going to make an impact on such a big company? I feel like I've figured out how to do that, but now the challenge is execution. We're much further along than I expected to be six months ago. It's a testament to Sun and its willingness to innovate, to change.
Further along with what?
Project Indiana. (See also Stephen O'Grady's analysis.) The goal is to do a few things around Solaris to make it more approachable and appealing to developers who would otherwise go to Linux. People hear "OpenSolaris" and think that it must be like "OpenSUSE." In other words, a community, binary distribution of a wider body of code. So we're trying to give the developers what they already expect from the Linux world: an OpenSolaris binary distribution.
This involves the following:
- Build developer mindshare around Solaris. We're living in a world where there are more people familiar with Linux than with Solaris. We want to position Solaris as a great alternative to Linux which can do things it can do, and some that it can't.
- Make Solaris more readily comprehensible to Linux-aware developers. This is partly just changing the nomenclature around Solaris to make it more familiar to Linux users (e.g., calling it a "distribution"). But it's more than just casting it as a "Linux distribution" of sorts, but also bringing to the forefront the differentiating features of Solaris.
How do you compete with Linux? Specifically, how do you manage the application certification problem?
There are actually more applications certified to run on Solaris than Red Hat Enterprise Linux today. So the application line is a bit of a red-herring argument.
It is true, however, that Solaris is behind in terms of desktop applications certified to run on it, but that has more to do with its history as a server operating system. We don't have any particular designs to make Solaris the industry's primary desktop operating system, but we need to have it run on developers' desktops because that's the gateway to their use of Solaris throughout their enterprise (e.g., in evaluating new applications to run within their enterprises).
Sun is actively involved in a wide range of open-source projects. Is there a unifying theme to Sun's software strategy and open-source activities?
We're a systems company. Customers want integrated solutions, but this has traditionally meant "take it or leave it" for customers. What Sun is doing at every level of the stack now is to componentize its software stacks so that customers can take the entire stack, or can replace components of the stock as they wish. We drive value to our stack by optimizing the pieces to work together, but want customers to have the choice to replace our component parts with third-party components that they may prefer.
I once heard you say "Linux is Linux is Linux." Is "Linux is Linux is Solaris" also true? Meaning, is there really much of a difference?
That statement is largely true because all Linux distributions pull from the same tree. Solaris, however, is different on several different levels.
First, technology. Sun is and always has been an innovative company. Recently we've introduced innovations like ZFS and Zones (lightweight virtualization technology).
But we also differentiate in non-technology ways. The fact that Sun has been in the enterprise for several decades means that we understand our customers very well. We have the experience and expertise to provide global, true enterprise-class support. We also know how to patch and upgrade without breaking backward compatibility. These are compelling differentiators that no Linux vendor can match.
If Sun succeeds in its comeback plan, a lot of credit will go to Ian. He clearly believes in the company's direction, as does Jonathan Schwartz, Sun's CEO. Get a few evangelists around the table and next thing you know the company will believe, too. From there it's not hard to see their enthusiasm carrying over into the wider IT industry.