Jonathan Schwartz: Oracle bungled its chance at mobile Java
Oracle could have innovated with Java instead of litigated, says CareZone co-founder and Sun's former CEO. Also: Why Amazon won at cloud computing, "Intel Inside" was a blip, and the Mac is back.
Instead of leading 30,000 employees at a beleaguered Sun Microsystems, Jonathan Schwartz is now leading just a dozen at his.
But Schwartz remains the same. True to the provocateur culture that helped keep Sun in the headlines despite a relatively small advertising budget, Schwartz clearly relishes holding forth about the trends that will separate the computing industry's winners and losers.
Among some opinions Schwartz shared in a recent interview: that Macs will once again seriously compete with Windows for PC market share, that Oracle lost a chance to innovate rather than just litigate in the mobile software market, and that the fortunes of Linux and Amazon underscore his two biggest regrets at Sun.
Sun acquired Schwartz's earlier startup, Lighthouse Design, in 1996. He rose through Sun's ranks, leading the server maker's acquisitions, its ill-fated partnership with Enron, its software group, and eventually the entire company. After a decade failing to recover its dot-com glory, he then went quiet after stepping down with a .
Oracle saw Sun's Java software differently than Schwartz. Java combines a programming language, programming tools, and accompanying software into a foundation that lets Java programs run across many devices, and Sun had led an alliance of companies that successfully spread it across Internet servers and mobile phones.
Google almost became a Java ally when it needed something to help its Android mobile operating system span many devices. Instead, Google plucked what it liked from Java but left behind the brand name and its promise of app compatibility. Schwartz wasn't happy but . After acquiring Sun, though, Oracle sued Google for copyright and patent infringement. When , Schwartz than a pushover.
The fast-talking ponytailed capitalist do-gooder from San Francisco now is promoting CareZone, a socially networked information-sharing utility -- a subscription service Schwartz hopes will "help you connect with your loved ones, your siblings, your spouse, to get organized and stay connected with the stuff that matters." He chatted with CNET News' Stephen Shankland in a restaurant not far from the San Francisco home where of his dozen employees now work, and here's an edited transcript of the conversation.
With CareZone, why not just use Google Docs or some other service and share it with your babysitter, your nurse, your hospice program?
Schwartz: People keep information in Dropbox, they keep it in Evernote, in Google Docs, in Microsoft Excel, they keep it in e-mail. The point is they keep it all over the place. What we do is provide you with a purpose-built environment oriented for this. If you want to keep a journal, a set of private notes or public notes, documents, medicines, contacts, and have them securely shared in an ad-free environment that's totally private and walled off, that's what we provide.
Ad-free and walled off? So do you think you'll be more impervious to some of the unpleasant side effects of running an online business, like advertising?
Schwartz: Inevitably if you're in the world of social media, you have this conundrum to deal with, which is that privacy is anti-revenue. The more private Facebook is, the lower their revenue it gets, because Facebook gets money by selling access to you to advertisers. That's neither good nor bad, but it's a different model than I want to pursue with my parents or with my children. I want that to be in a private place where I control who's got access.
When you started CareZone, it's a very different environment from when you started Lighthouse Design or when you were at Sun. How much easier is it today to do a startup than five, ten, or fifteen years ago because of Amazon Web Services and SalesForce.com and things like that?
Schwartz: Immeasurably easier. Our single largest expense second to payroll is visiting the Apple store. I can't think of any other line item that even comes close. We use all the latest and greatest available services from all the best vendors you can think of. We don't have to pay for and take delivery of a data center, because we can buy all the services off the Web.
What's the difference running a business when you're pitching products to consumers instead of big businesses?
Schwartz: A third of our business [at Sun] was financial services. The biggest difference is at CareZone I have perfect data. I have really smart people who build very measurable things. Every morning I wake up and I get my automated feed that happened yesterday. Every interaction, every registration, every invite. At Sun, it would be great if every 90 days I got the same amount of information. There's so little data compared to what I have. The infrastructure at CareZone is built on the assumption that we'll be measuring it all the time. When we make decisions -- does this work or that work -- I get feedback. I can tell within 48 hours whether it actually worked.
One thing you did at Sun was the [Solaris] "release train," with quarterly updates and a subscription model. Are we moving to a world where software is continuously updated and nobody ever knows what version they're running? Where does that work and where does that not work?
Schwartz: For me, there's really a very fine line between media companies and technology companies. A technology company that's defined by the last product they built, that has to rely upon that for all future revenue, is like a newspaper company that has published a bunch of stories and now hopes they can sell them for the next 12 months. It just doesn't happen. Every day there's new competition, there are new opportunities. For the consumer Web, I don't think that model can work.
The model that does work is inevitably going to be a blend of just go update the Web page [and] we'll give you a runtime that'll go on your iPhone via the App Store or your Mac or your new tablet, and we'll keep it up to date for you. You'll have the ability to run off the grid so if you're in the subway tunnel or at your grandmother who doesn't have Internet connectivity, you can still get access to CareZone.
I think that will work for the vast majority of consumer applications. Are you going to update the software on your [Airbus] A380 all the time, every day? I sincerely doubt it. They have very different standard of quality. In a consumer world, they just want to know that it's good enough. They want to know that some things are flawless -- security and privacy being the two most important to us. But we're not running Bank of America. That's what Sun and Oracle are all about. They have to worry about the 911 network never being down.
You were one of the prominent people involved in the Google-Oracle lawsuit [called by Google to testify]. For now, that case largely fell in Google's favor.
Schwartz: I'm not going to opine on who was right and who was wrong. It wouldn't have mattered who called me to testify, my testimony would have been the same. All I knew is [at Sun] we wanted everyone to pick up Java out of self-interest. If they did, we would have a greater shot at making money in other related places.
You can't stand on both sides of the open-source dividing line. You can't say, "I want everyone to go pick up my software and use it around the world -- except for the people who are going to make money. I want them to come pay me first." I thought the judge was very clever. Good thing he was also a programmer so he could go and implement RangeCheck [a function in Java that's also used in Android that was a source of contention in the case]. I thought the outcome was fair and right.
Is Java is stronger or weaker as a result of Google and Android? [Android uses many Java functions and the general Java approach.]
Schwartz: That's not my world anymore. I would like to see Oracle invest more in Java. I would like to see them doing more with Java running on handsets. I think there was a great opportunity for Oracle to get out there and lead the market.
There's no doubt at all that given the success of Java, the pervasive adoption around the world coupled with one of the most successful companies on Earth, that they could have delivered amazingly. It was their choice not to. To disinvest in all the work we did at Sun -- what's going on with Java ME [Micro Edition, for phones and other smaller devices], JavaFX [a variant of Java designed to ease compatibility problems] -- there's a whole bunch of stuff that's ripe to bring to the marketplace. Sure, you can say it won't work, but it still was a valuable business.
It was a choice. I'm not going to cast aspersions on Oracle. They have to decide what's best for their shareholders. They spent an awful lot of time and energy litigating. If they'd spent that energy building a great handset or a great developer platform for mobile devices, there could have been a different outcome.
It's what you faced at Sun when Google went its own way without taking a Java license: litigate or innovate.
Schwartz: It's exactly that. But you have to remember that just because you don't want something to happen doesn't mean it's not going to happen. One day somebody will write an interesting analysis of the last 10 years of that company's life, but the single biggest threat at Sun was its single biggest opportunity. Linus [Torvalds, founder and leader of Linux, which competes with Sun's Solaris version of Unix] wanted access to the Solaris source code. At the time, management said no, and it created its own [the OpenSolaris project]. Imagine what would have happened if Sun had said, "We're going to make the code available under this license. Now let a thousand flowers bloom." [Linux seller] Red Hat is now worth more than Sun was.
When you were launching CareZone, did you consider whether the patent system favors the incumbents who have big portfolios? Did you go out and look at who had patents on something that might bear what CareZone is doing?
Because of willful infringement? [A court finding that patent infringement was willful -- in other words done knowingly -- means that the infringer must pay triple the monetary damages than if infringement wasn't willful.]
Schwartz: Absolutely. I don't think the patent system favors incumbents. I think the patent system favors non-operating entities [also known as non-practicing entities, or NPEs, which buy patents and license them, sometimes under legal pressure, but which don't have a technology business that might give another company a patent infringement counterclaim]. That's what's really scary, because now you've got somebody who's got nothing to lose. [former Microsoft chief technology officer and co-founder of patent licensing firm Intellectual Ventures] is a very, very, very smart man. He may be the wealthiest man on Earth when all is said and done. Congratulations on arbitraging the patent system.
Do you think with Google and some of the recent decisions in the Apple patent lawsuits that some of the wind will go out of the sails of patent litigation? There's been a lot of litigation, and it doesn't seem to have really moved the needle.
Schwartz: If you look at the server market, there's one dominant patent holder who wields it like a cudgel.
International Business Machines?
Schwartz: IBM. When they could, they'd bop you on the head and say you're stepping all over our patents, you need to send us a bunch of money. Reams and reams of companies did it. Then IBM ran up against a wall, because the next big set of people they could sue would be their customers. So basically they quiesced that activity. The places that did not quiesce that activity were the carriers and the mobile handset manufacturers. The carriers were vicious and voracious, as were the standards bodies. It was never really a factor in enterprise IT. I mean, it was in the IBM days, but their core business wasn't suing people. For a lot of guys in the mobile handset business, that was just standard operating procedure.
How successful do I think they'll be? Not particularly very successful. I think they will create lots of heat and not a lot of light, and I don't think it'll be a line item on anybody's income statement. But it's a chip to play. It's your ante. You've got to get out there and sue the crap out of everybody else out there.
We have a policy [at CareZone]. You will never read a patent. Never. Only your own.
If my memory serves me, your name is on a patent [application] for charging per-person, per-year subscription payments for software. Are you going to go after Google now that they charge $50 per person per year for Google Apps? That could be a nice revenue stream for you.
Schwartz: We revoked that patent. I don't think it passed the red-face test. Patent litigation is not how I want to make a living. It's not the legacy I want to pass on to my children.
When you buy Microsoft Office, you buy a perpetual license. What do you think about people who still sell a perpetual license?
Schwartz: I think there a couple reasons that gets harder and harder. One of them is pricing. When you're a CIO, you want to get down to how much is supporting going to cost? If the vendor has done the work for you, they're more likely to get your business. But if [the vendor offers] some big mushy thing plus x percent plus y percent -- how do you know? Companies like SalesForce and Google say, "Here's the price. Add it up." It makes it a lot easier to make a decision, as opposed to having to meet with the sales director, go golfing, and hang out on somebody's yacht.
Adobe has just begun making this transition with its Creative Suite, charging $50 per month if you sign up for a year to the
Schwartz: They're a really really smart company, and they are unique in that although they have a tremendous enterprise revenue stream, they tend to sell to the users -- the designers, at least on the creative side.
[Price] transparency is your friend. If you're competitive, you want everybody to know how competitive you are. It's only when you have something to hide that you want to hang out in the background and try to avoid letting people know what your product prices are.
At Sun, you were into utility [pay-as-you-go pricing for server capacity with the Sun Grid]. Why did that happen at Amazon, with Amazon Web Services, and not at Sun?
Schwartz: There are probably two things I regret. One is I didn't GPL [in other words, release Solaris' Zettabyte File System under the GNU General Public License, the prominent free and open-source software license that governs the Linux kernel. Sun chose its own incompatible license instead so ZFS couldn't be incorporated into Linux.]
So it wasn't in Linux, where it would be relevant, but instead walled off in a little Solaris domain where it was in a little niche?
Schwartz: Yes. That's one. The other is some of the choices I made when we started delivering that product per CPU-hour. We made a classic mistake. I made the mistake. We went after some of the largest customers in the world. Amazon went after the smallest customers in the world.
Schwartz: Right. And developers. Now they are the global leader. I'm not even sure who's second, because I'm not sure it matters. You have IBM and HP and a whole bunch of other companies saying, "Ooh, wow, let's hang out with Morgan Stanley and sell them a computing cloud." Wrong answer. Now you have nine months of a procurement process, three months with the security department, four months with the IT group that's in charge of making the decision -- and Amazon just got another 200,000 customers.
We had that instinct in every other part of our business. We didn't in this, and it was a mistake. It was a big mistake.
What phone do you have, and what computer do you have?
Schwartz: I have an iPhone 4S because there is no iPhone 5. I have a MacBook Air. I work in a company that's 100 percent MacBook.
Why is that?
Schwartz: Because they enable people to get their jobs done. It's just easier to get work done. I want to spend time communicating with people, creating content, interacting with customers. I don't want spend it futzing around with some obscure bug in some other operating system or dealing with some special new antivirus protection that that company's trying to shove down my throat.
That's definitely an inspiration for most companies these days. You want to be an invisible utility.
Do you think the PC market share will open up again. Will it be like the 1980s again when Apple and Macs had a chance against Microsoft?
Schwartz: There's no question in my mind that market will heat up all over again... If you go to an employee and say, "You've got two choices. I'll give you an iPad or I'll give you that last-generation x86 box," they'll take an iPad. The price of an iPad is $400 or $500, and the price of IT and connectivity is plummeting. I have always said enterprises have to follow their employees. Why did PCs end up in the enterprise? Because employees wanted to take their work home or wanted to bring things from home into work, and you couldn't do that with your 3270 terminal [a "dumb terminal" connected to a central server]. It's all well and good if you control your employee domain, but the moment you start opening up your services to consumers, you don't control the domain anymore, which means you have to prioritize dealing with heterogeneity.
You positioned the choice as between an x86 machine and an iPad, not between an x86 machine and a MacBook.
Schwartz: I think they're related, because they'll all be the same user experience. It already is going in that direction. I think the discussion over chip architecture is inconsequential, because no customer I know wants to say, "Oh, look at the new iChip in my device." They don't care. They care about the user experience.
So Intel Inside [Intel's marketing campaign to promote computers using its processors] was a blip?
Schwartz: I think it was a blip.
From a day when consumers cared about what microprocessors were inside their computers.
Schwartz: I'm not sure they ever really cared. ARM [an Intel rival strong in mobile computing] is probably proof in the pudding. What [consumers] cared about was getting their work done. Intel happened to have a near-monopoly on most of those devices, so if they stuck the brand on the side, that was very clever follow-on marketing.
End-user experiences are the singular thing that matter. End-user experiences happen to be powered by microprocessors, but end-user experiences are the thing that's going to blow up the PC market -- if it hasn't already been blown up.
The question for some of the big PC vendors is, I see some of their ads in the paper advertising their new thin this or that. But what's the compelling user experience?
What's it do for me?
Schwartz: They don't have an answer for that.
They haven't had to do that marketing. They haven't had to carry that water for years and years. That's what Microsoft did.
Schwartz: And they didn't have the gross margins either. Now they're caught between a rock and a hard place. Microsoft is increasingly going to get into the hardware space. So where do [PC makers] have the expertise or the gross margins to develop the expertise?