X

Can software pull Sun out of its funk?

Sun's new software czar Jonathan Schwartz is brimming with ideas--not the least of them being a plan to launch the company's next assault on Microsoft. One problem: He doesn't have much margin for error.

Stephen Shankland Former Principal Writer
Stephen Shankland worked at CNET from 1998 to 2024 and wrote about processors, digital photography, AI, quantum computing, computer science, materials science, supercomputers, drones, browsers, 3D printing, USB, and new computing technology in general. He has a soft spot in his heart for standards groups and I/O interfaces. His first big scoop was about radioactive cat poop.
Expertise Processors, semiconductors, web browsers, quantum computing, supercomputers, AI, 3D printing, drones, computer science, physics, programming, materials science, USB, UWB, Android, digital photography, science. Credentials
  • Shankland covered the tech industry for more than 25 years and was a science writer for five years before that. He has deep expertise in microprocessors, digital photography, computer hardware and software, internet standards, web technology, and more.
Stephen Shankland
7 min read
Jonathan Schwartz took a job one month ago that most ambitious high-tech field executives would not relish: running a software business forbidden from ever standing on its own.

Within the confines of Sun Microsystems, where he is executive vice president of software, Schwartz is nonetheless brimming with ideas--not the least of them hatching plans for the company's next assault on arch-rival Microsoft.

That's not all. He's talking about selling Linux-based desktop computers for specific jobs and backing the MySQL open-source database software for certain tasks. Schwartz is also figuring out Sun's new "N1" plan to let administrators manage thousands of servers and storage systems as a single computing resource.

A fast-talking 36-year-old with a ponytail, Schwartz has made a rapid rise within the ranks since Sun acquired his software company, Lighthouse Design, in 1996. He ran Sun's $400 million investment fund--a program that expired when the Internet bubble burst. In 2000, he took over strategy and planning, where he led acquisitions work and became Sun's top dog in the Liberty program designed to outflank Microsoft's Passport online identity system.

Despite his new role as Sun's software czar, Schwartz faithfully sticks to the party line, arguing that the company can reap bigger dividends by combining software and hardware than by subordinating one to the other.

Microsoft is a great technology company. But their monopoly position has shielded them from the realities of the marketplace.
"Everybody keeps wanting to prove that we're not a software company. Well, stop proving. We're not," Schwartz said in an interview with CNET News.com. "I don't want my staff to grow an independent business that has no value to Sun. We're not a hardware company, we're not a software company, we're a systems company."

The subordinate role of software at Sun stands in stark contrast to Sun's two biggest competitors. Like Sun, IBM sells servers, but Big Blue also has independent software businesses for databases and e-commerce programs that run on other companies' servers. And of course Microsoft relies almost entirely on third-party vendors for hardware.

Software remains a low priority--even an afterthought--at Sun, according to some analysts who believe it could help the company remain a step ahead of the Microsoft-Intel-Dell juggernaut. Schwartz has other ideas. He recently spoke with CNET News.com about this and other issues connected with his role in shaping Sun's new strategy.

Q: There's been an ongoing debate about whether Microsoft is putting Java in their browser.
A: Microsoft is doing whatever it can to confuse developers in the marketplace, to continue to promote the idea that every platform except Windows is unstable on Windows. That's why the federal government brought suit against Microsoft. For example, just before MacWorld (the trade show), guess what Microsoft said? "We're not sure the Macintosh is viable." That just injects enough uncertainty into the system that people tip their purchase decisions.

StarOffice is available free from OpenOffice or at a nominal price if you want to deploy that in an enterprise. Are we going to be building that? Yes. Do we believe there's a healthy market opportunity to deliver a Linux client and do call centers, payment processing centers, reservation systems and factory floor plants? Absolutely. You've already seen us tip our hand. We've delivered the office suite that's necessary. The Gnome community has delivered the user environment. All we need is a browser to make sure we round out the trio.

Microsoft is a great technology company. But their monopoly position has shielded them from the realities of the marketplace. Among them: Customers don't want to pay $500 for an office suite. They don't want to pay another $500 for a Windowing environment. They don't want to have their customers stolen by Passport. They don't want to have their media locked up in Windows Media format.

Are you trying to get in front of the train in Web services?
We're the locomotive. Think about it. They're all written to Java. The only issue there is the adoption of XML. Were we late to the game? Yeah. But now we're driving it. JAX-RPC is a standard we created. Am I worried about SOAP and UDDI? Not in the least.

N1 is under your purview. How much of N1 is a software design?
N1 is all software. The idea behind N1 is simple. If you run lots and lots of low-end systems, you need to worry about provisioning and management that you didn't (have to worry about) when all you ran was one E15K (Sun's top-end 72-processor server). All we're doing is taking the technology in an E15K and repurposing it for 2,000 Netra T1s and x86 blades.

When is the vision of N1 coming to reality, when you can manage an entire data center the way you manage a single server today?
It'll take years to deliver. We've got 25,000 systems under remote management and monitoring. We are about to ship iChange in the next release of Solaris. So there are elements of it today. N1 is very much the banner under which this stuff will be aligned.

You talk about lots of software "sedimenting" to become part of the operating system. Do you think databases are going to become bundled as a standard feature of the operating system?
Stay tuned.

What does that mean? You're interested?
Oh, sure. I think it's a huge competitive opportunity.

How would Sun go about having a significant presence in the database market?
Have you noticed we already ship MySQL on Linux and Solaris? What I've heard back from customers is that (IBM's database) DB2 is outrageously expensive and (IBM's e-commerce software) WebSphere is fifty thousand bucks (per) CPU. Our value proposition to them is, "Why on earth did you pay fifty-thousand bucks a CPU for WebSphere when you can get a free Application Server 7.0 from Sun running on Solaris and Linux--and by the way, we'll give you a database as well? Why do you bother with DB2 on Linux? Why don't you just run MySQL? It's cheaper, faster and more stable..."

I was just at Google, and Google's an all-MySQL shop. Why did they do it? Because they looked at DB2 and it was expensive and it didn't offer any added value.

I haven't noticed a very rich independent software community living on top of MySQL the way one has on Oracle.
Go check it out. I don't think they have to be in the (high-end) data center business to be interesting to us.

Are you concerned that Oracle, one of Sun's biggest partners, is pitching a competitor?
If you want to use Oracle, by all means, please do. I think the issue Oracle faces is they're trying desperately to embrace Linux, and Oracle's "unbreakable Linux" (pitch) certainly makes a statement. My retort would be unbreakable MySQL.

A lot of the big bets IBM has been making--Java, Web services, Linux--are things that span their multiple hardware environments. Does that worry you?
Tell me what their unification is.

I think we have been individually innovative, but not as innovative at a systems level.
You can run software you wrote for WebSphere on one system or on another. Software you compiled for Linux, you can shift it around more easily than you could for AIX versus z/OS. You could write Web sites or other business transaction systems written with Web services standards and move them more easily from one IBM system to another.

It doesn't worry me because the cost of porting is negligible compared to the cost of qualifying. You can run Linux on a mainframe, but guess what? You're going to emulate it, then you're going to run it on a custom emulation virtual machine on a custom z/OS on a custom microprocessor that I can't even name.

What's the biggest challenge you face right now?
Having 5,000 people who are all incredibly passionate about what they've been doing for the last two years.

But?
Needing to get 5,000 people passionate about what my staff wants them to do for the next two to three years.

If the focus is on integrated hardware and software, what happens to the versions of Sun One software for HP-UX, Windows, for example?
So long as customers want them, we're going to keep delivering them.

What percentage go onto non-Solaris computers?
The high was probably in the mid-20s, and now it's probably in the low teens at best.

Do you like Apache (the Web server software that competes with Sun's own package)?
We ship more Apache than any other company. It's part of Solaris.

Is it a better idea to throw your weight behind Apache than the (Sun ONE) Web server?
While it's still a very healthy business and customers are willing to pay us for the value addition, why not? We have no problem embracing both.

What changes do you expect under the new Schwartz era in the software business at Sun?
A much stronger degree of alignment and integration among all our products. I think we have been individually innovative, but not as innovative at a systems level. We have an extraordinary breadth of assets, unparalleled in the industry, even by Microsoft. We go way down below where they are, we go way up above where they are. The challenge for us is to go deliver those as solutions (rather than isolated packages).

For example?
For example, when we talk about an identity solution (software to track computer users' identities such as login names), it's not integrated just with our directory, it's with clustering so we have a highly available network identity solution--with JavaCard so we have microprocessor smart card element of a network identity solution, with the Portal Server so we have a network identity enabled single sign-on portal solution.