With his management team reshuffled, Suns president and chief operating officer Jonathan Schwartz sat down again with eWEEKs Steve Gillmor to discuss Suns contributions to the open-source community, the Sun-Microsoft deals impact on DRM, identity management and visual development tools, and the strategic role of Suns auto-update technology in the emerging real-time enterprise platform.
Steve Ballmer told eWEEK editors the deal had nothing to do with Linux, that it was only about Solaris. Doesnt that differ from your perspective?
No, not in the least. I think what hes focused on is ensuring that Microsoft is paid for the protocols that are licensed when delivered in commercial products. But clearly, great portions of Suns products are built with the contributions made by the open-source community—for example, StarOffice.
Should we license technology from Microsoft, it would clearly advance the interests, for example, of the Java Desktop System, and if thats running on Linux, then that obviously helps the underlying Linux community as well as the overall growth and viability of open-source technology.
Just remember, there are a great number of non-open source technologies in the open-source community that add value to the open-source community. Ill give you one very good example: RealPlayer. RealPlayers not open-source, but its availability certainly enhances the value of our Java Desktop System. Macromedia Flash—not open-source—clearly enhances the value of the Java Desktop System.
Adobe Acrobat Reader—same idea. One of the values that Sun has represented for an awful long time is our ability to synthesize work from the open-source community with work from the standards community, and ultimately drive both to deliver value to our customers.
On a practical level, if Steves saying these licenses are only for Solaris, how do you release the same code across all three platforms?
StarOffice, as an example, runs on almost all platforms shipping today. There is work done in the open-source community to advance StarOffice, but certainly Sun stands behind and indemnifies StarOffice. To the extent that we license protocols from Microsoft, we would be including them in StarOffice—and not obviously distributing them free of charge—just as we do today with RealPlayer.
We incorporate RealPlayer into the Java Desktop System, but we dont go delivering it out into the GNOME community to add value to.
How do you leverage the Unix licenses you already have as they relate to the Microsoft relationship?
There is nothing that precludes us from taking the protocols we license from Microsoft and incorporating them into our products. Now, where those products run is up to Sun. So, if we take a license from Microsoft, theres nothing that precludes us from incorporating that technology into our Java Desktop System.
Again, we are competitors, and it is every bit in our interest to see how many Windows users we can migrate into the Java Desktop System, as it would be for Microsoft to see how many Java developers they can try to sway into the .Net world.
Beyond the Code
In our previous conversation, I had a little trouble understanding what the value to the open-source community would be.
Well, whats the value of RealPlayer to the open-source community? It validates the work from the open-source community. Whats the value of StarOffice to the open-source community? It validates the work of the open-source community.
The Java Desktop System validates Linux as a viable enterprise alternative to the Windows environment. Adding value to the open-source community doesnt always mean adding code—it means adding marketing awards, endorsements, a seal of approval, developers [and] programs.
Its a naive analysis of the open-source community that says its all about forking over source code. Its not—its about building community, about making investments in marketing, in developing technologies that run on, with and through the open-source community. We have a very long history of working with the open-source community.
Despite some of our peers in the industry who hire people with titles like evangelist, our folks have titles like developer and architect, and they go work with the open-source community to build technologies and solutions that solve customers problems.
I would point back to the Java Desktop System as evidence of the work weve done with the GNOME community, the Mozilla community and the Linux community to really bring products to market that dont just add more lumps of source code into the source tree but deliver value to customers so that they want to inject money into an ecosystem and make it self-sustaining and profitable.
How would you compare the Java Desktop System to what Novell is putting together, given common elements such as Evolution?
Certainly it bears a relation in that were both hoping to go be competitive against Microsoft. The work weve done has been more than simply assembly, though. Were one of the principal contributors to the GNOME project, to Mozilla, obviously to StarOffice—and we believe in not simply assembling technologies but really commercializing them.
Novells participation in the market is a good thing, because it validates the market as creating an opportunity for more than simply one company. So, I welcome the competition. To us, its really emblematic of the nature of the relationship we have with Microsoft, which is a deeply held belief that a rising tide lifts all boats, and that interoperability between Sun and Microsoft grows the overall market for both of our products rather than advantages one company versus another.
Choosing Sides with DRM
Youve indicated that you would look for a unification of Microsofts and Suns DRM strategies. Is that what you really mean?
I firmly believe that identity is the cornerstone of all network services. You cant expect to build a network service unless you know to whom youre delivering it and what they have access to.
The focus on identity is really a focus on a core area of our business where, through our Java Enterprise System and the directory and identity technologies that we have built to run that, we currently have market-leading products.
The acquisition of Waveset [Technologies] last year has really put us at the front of the marketplace for provisioning systems. So, its a natural extension in the market: how identity services may or may not necessarily [play] into services and service access but into content types, into the DRM standards through Microsoft Office—wanting to ensure we have interoperability in StarOffice, ensuring that the DRM that is flowing through Windows Media we can make available through the media technologies that we will be delivering.
There are those who say Apples DRM strategy is less onerous and provides an alternative to the Microsoft strategy.
There will be a diversity of DRM solutions in the marketplace. There will be no one size fits all. Any illusion that either Sony or Philips or Microsoft or Apple or anyone else is going to have the keys to the kingdom through DRM is fundamentally naive.
But all that said, if you look at the market for music downloaded on mobile handsets, you measure it in the billions of dollars. If you look at the market for music downloaded on desktop computers, its measured comparatively in pennies. Now, why is that?
Its principally because the mobile carriers have a much stronger grasp of identity in the infrastructure necessary to monetize content. Independent of their DRM solution, theyve made it convenient and theyve made it authenticated—those are really the keys to monetizing content—as opposed to perhaps a more accepted focus on the specifics of the technology within the content.
Visual Studio Meets Java
Many analysts have focused on Visual Studio as potentially being enhanced to work across both .Net and Java. Do you see Java Studio Creator being similarly impacted?
We certainly hope and aspire to build better interoperability between Visual Studio and Java Studio Creator, and theres a common philosophy to how you build service-oriented applications delivered through both of those technologies. The delivery of byte codes that run on an equivalent of a virtual machine is common to both. Theres ample opportunity for us to collaborate there as well as realistically develop the others developer bases.
We both believe that Java Studio Creator is a perfect range of platform for developers who are comfortable with the Visual Studio style of developing applications—who may have been otherwise intimidating by enterprise Java—and it certainly brings it down to an audience that allows us to begin migrating out to an increasingly large market opportunity. Again, interoperability is about growing the overall market and not just picking off one anothers customers.
But you can already use Web services to interoperate. Whats the secret sauce here?
You can use Web services to interoperate, but the problem has been that you need a PhD to figure it out. For the most part, the vast market wants a visual development tool that allows them to get access to the same technologies without having an advanced degree.
To suggest that Web services are easy for everyone—Im sorry, I dont buy that. There are probably 10 million developers in the world, and of those, theres a very, very small fraction that are Web services developers.
The majority of people who are content creators are people who carry around digital cameras and these devices for a networked world. You can move up from there to people who create static content—more complicated documents and spreadsheets—and establishing what were going after with the StarOffice world.
But as soon as you get into more interactive content like Visual Basic, or StarBasic for that matter, you begin adding complexity that oftentimes excludes people from being able to get their jobs done.
So, with interoperability and a focus on ease of use, were trying to use both StarOffice as well as Java Studio Creator to create a broader market opportunity and add interoperability to that mix. Its about growing the largest market possible, trying to help build the biggest tent atop all the developers in the world rather than forcing people to go make choices that may preclude their opportunities.
-Update and Real Time”>
You and I have had many conversations about auto-update and what implications that might have for client-to-client technologies. What is the auto-update value proposition?
I think you and I are highly aligned on this, but the model of computing that says you go to visit a Web site to go get content is antiquated. We should expect to be in a world that is not only far more notification- and event-oriented, but increasingly oriented around session-based content types instead of simply static content types.
Thats clearly an area of interest to us because we see so much technology in the world that is either out of date or out of security or just simply could be of higher value to its clients and users if it could be kept synchronized.
Were going back to pub-sub, and were moving away from a concept of Im going to go to a Web site and pick up content. The pull model of computing—the days are numbered. Were moving toward a model where we have the opportunity to keep the technologies that Sun delivers up-to-date and current, as well as keep users up to date and current with what is going on in the marketplace.
In the Microsoft model, thereve been increasingly blurred boundaries between the OS, Office and the server middleware layer. If you dont use the browser as the focus of the desktop, and you move to a pub-sub router as the focus, doesnt that change the equation for how bits of information travel around the network?
It certainly changes. I do not believe the browser is going to be the long-term focal point for network services. The network is going to be in the long run the focal point for network services. And moreover, authentication is going to be the focal point.
Im not a big believer in the idea that youre going to authenticate devices—I think you are going to authenticate users. And those users are going to want to have access to the latest and greatest information at whatever device to which they have authenticated.
That also suggests, though, that theres got to be a delineation between the kind of things you do with a PC and the kind of things you do with a mobile handset or an automobile seat back or a cable set top box. The diversity of clients out there is going to be such that a browser in some instances will be an ideal interaction environment—one where youre likely to be stable for long periods of time, tethered, terrestrial—and Im going to want to be reading a lot of content.
The browser has proven itself to be a particularly poor interaction model on a mobile handset, an even poorer one on a set top box, and I can guarantee you it will be absolutely horrible on an automobile dashboard or seat back. That implies that were going to have to get cleverer about ways of delivering information to the diversity of devices attached to the network.
Event notification and concepts like auto-update … Auto-update is not simply about delivering the latest and greatest patch to an operating system—its just as much about delivering the latest and greatest Steve Gillmor column to a 10 million reader base.
Or the software that delivers that column.
Absolutely. We certainly look at the evolution of the Java virtual machine on more than 60 percent of the PCs in the marketplace as opening a market opportunity for us. How we elect to take advantage of that opportunity is a topic for further discussion, but certainly the mechanism behind auto-update is one means of delivering functionality, not simply patches and updates, but actual applications that enable subscribers to that auto-update function to get more information, more technology and more service from Sun and our partners.
Broadest Possible Market
In his conversation with eWEEK, Ballmer suggested that Microsoft needs a lot of customers to succeed. Further, he suggested that IBM has lost credibility in the enterprise because their software doesnt get used much. Does this new events-based fabric, running via auto-update as the glue between OS, apps and middleware, affect the customer equation?
The fact that Red Hat has the Red Hat Network, that Sun will have a vast auto-update network not only built from Java but from our Solaris operating environment, and that Microsoft has the Windows Update function, suggests that IBM has been somewhat hung out to dry.
They were not a part of the [Sun-Microsoft] announcement, and I think they have left themselves exposed in their zeal to move to Linux. They have failed to recognize that they have simply allowed Red Hat to replace Microsoft as the provider of operating systems for IBM.
Now that Red Hat is increasingly competing against IBM as well as raising its prices to the point where customers are really beginning to express their dissatisfaction by seeking alternatives—both back to Microsoft as well as to Sun with our Solaris operating environment—suggests that it was a short-sighted decision that said they were going to abandon AIX for Intel and Opteron servers.
And by the way, the same goes for HP, who managed to disinvest and move away from HP-UX just as the volume market was beginning to appear on Intel systems and then ultimately evolved onto Opteron. In my mind, we had the strategic courage to commit to Solaris on Opteron as well as the courage to continue to propagate Java into the world to create the broadest network possible.
So, I am very convinced, with Steve, that he who has the most customers—and I would just add developers into that—is ultimately going to be the long-term winner. Companies that believe you can use press releases to curry favor come up short in the long run if they dont have technology in the network that allows them to reach the marketplace.
Do you think its an interesting coincidence that Microsoft released some of its auto-update technology to open source?
Nothing that Microsoft does is ever a coincidence.
This is the first time that theyve come up with a license similar to the IBM or Apache licenses.
Microsoft is filled with some of the smartest software developers in the world. Fundamentally, we have something deeply held in common between Sun and Microsoft: We both value intellectual property. And we also both value the innovations from extremely creative individuals. And the market isnt shrinking from network services or network infrastructure.
The focus of our work with Microsoft was to create the broadest market possible for both of our products. Our belief is as the market grows, both of us will have more opportunity. We will continue to compete by providing the Java Desktop System as an alternative to the Windows desktop system. Well continue competing to provide Solaris as an alternative to the Windows operating system, and the Java enterprise stack to the middleware stack that Microsoft delivers.
At the end of the day, thats great for customers. The fact that were committed to interoperability means either choice is a safe choice. Were very bullish on the future of the network and very bullish on the future of intellectual property in open source as well as in open standards to continue to drive that opportunity.