How is Sun going to make money from Java?
A candid interview with Sun Chief Science Officer John Gage
John Gage speaks his mind. Who else would call Apple's top executives "thugs" and admit that his own company's corporate structure is "wrong?" Here he also sounds off about how Sun is really going to profit from Java and why the rumored Apple/Newton acquisition is not going to happen. (1,700 words)
In the early days of Java, a reporter's attempt to confirm an utterance by John Gage about the future of the programming language was met with a sigh at Sun Microsystems Inc.'s public relations department.
"Oh that's just typical John," the representative said of the company's chief science officer. "He's always a few years ahead of the rest of us."
Staying ahead is Gage's job. Since he joined the Menlo Park, California-based company in 1982 as one of its original members, Gage has traveled the world finding promising technologies and people that might help position Sun for the future.
Gage and Sun co-founder Bill Joy, for instance, planted the seeds in the mid-1980s for Sun's forthcoming SunScreen SKIP E+ encryption technology (see Resources below) when, while visiting the Soviet Union, they met an engineer in a restaurant amid the clamor of Russian weddings. Gage was also the front man for Sun's involvement in Malaysia's Multimedia Super Corridor, a trade zone project that won't fully bloom until after the turn of the century.
As an advisor to the key decision makers at Sun, Gage is privy to company strategies and at times speaks freely on issues other officials won't touch, such as Sun's flawed corporate structure and its failed bid for Apple Computer Inc. Gage did just that with Rob Guth last week while in Tokyo for the NetWorld+Interop trade show. Below are some excerpts.
Guth: How does Sun have to evolve over the next few years? Will the company have to make some structural changes?
Gage: The structural form is wrong. We divided a long time ago -- in really not a very thoughtful way -- into a hardware part and a software part. That division doesn't make sense to me ... it makes you think incorrectly about what you are doing: "Does this person belong in the software part or the hardware part?" Instead of thinking "is this person going to make networking work whether or not it's the hardware or the software [company]?"
That's cost us a lot. We probably hired twice as many engineers in the software area as we need but you can't get rid of them. It just worked out badly.
The main hunk of Sun will be focused in three or four years pretty much where it is today: powerful file servers, very high speed networking, very reliable machines running Unix. Then the other part, the part that brings the seeds of change -- the Java stuff -- [will be] very small embedded stuff getting into automobiles, getting into appliances and that's new software, different expertise.
The new direction that Java has given us is the definition of how computing works in every area of computing. In the past we were not a presence in embedded stuff -- who cared what Sun thought about embedded [devices], they're not making chips that run in cars ... Now because of Java we are at the heart of the API definitions. Pick 20 areas of computing -- the Java API stuff -- and though in many cases we're not leading the group we are part of the group. In those areas we are suddenly back in the mainstream of development which means we recruit the bright people, they come to Sun. They will see that what they do turns into products.
Why Java is better than Solaris
Guth: How will you move into that new area?
Gage: We might end up splitting up. It is possible that we might make it a separate division ... we currently have a hardware company and a software company but the hardware company makes desktops which compete heavily with Windows NT ... so that's not a high margin business.
NT right now is moderately reliable but not very reliable ... it's easy for people to use but in the long term it's dead. It can't be made secure. Same with Unix. They're both big, 20 million lines of code, written in C and C++, you can't maintain them -- you have to wait for the company to fix a bug. It's just too slow. And it can't scale very well. You can't run Solaris in your wrist watch. You can't run NT in your wrist watch. You can run Java Virtual Machine in your wrist watch.
Guth: Will a group involved in this new hardware area grow from within the existing hardware group?
Gage: Let's add in people who are more consumer[-oriented] in their thinking. You can't make an engineer who is used to building a machine that costs $50,000 ... think the same way as somebody that builds $500 devices. They just don't think the same way. They can't be retrained. The cheapo guy is constantly thinking about cost reduction, saving a gram of plastic...
What we'll probably end up doing is form close partnerships with some major consumer electronics companies that have that kind of way of thinking and Japan is the perfect place.
Guth: How is Sun going to make money from this area?
Gage: We'll have partnerships on the underlying technology that makes it work ... we're not really designed to take advantage solo of this new arrival of these small devices. The effect of Java for Sun has been primarily to double our stock price because of the public awareness of Sun. People know Java more than they know Sun.
As far as bringing revenue to our bottom line I think the fundamental impact for Sun's future that Java has had is to bring a new wave of smart people into Sun.
Apple acquisition "too messy"
Guth: What about rumors of an acquisition of Apple's Newton division?
Gage: It's too messy. The killer in acquisitions is when you hire somebody who looks like you. Both have a group of people that do the same thing -- there's a graphics group at the Newton operation and there's a graphics group at Sun. So who gets fired? We thought at one point of buying SGI [Silicon Graphics Inc.]. They have some very nice, high-end file server machines. They have two or three designers who are just wonderful. Just world class ... but if we bought them we've got to go through all of their marketing people, fire them. We have got to go through all the people in comfortable job descriptions and figure out who do we keep and who do we throw away? It's just horrible. It's messy.
It's far better to buy something that's not like you. The Newton group is enough like us that it would be difficult.
Guth: Is the consensus now that Apple or the Newton group is not an interest anymore?
Gage: It's a difficult marriage question. We have a lot to do -- don't rule it out. If Apple really founders -- I mean they're foundering right now. The moment [chairman and CEO Gil] Amelio and [executive vice president] Ellen Hancock showed up, all the smart people quit. You could see it from the moment they walked in the door. They didn't understand how the computers worked; they didn't understand the system and people quit. And then Jobs is now there and Jobs will get them out instantly if he comes into power. He just can't tolerate people like that.
Essentially they're thugs. Their job is to cut costs and try to make the company survive financially, but they have no appreciation for what the heart of the company is. So all of the people that made Apple what it is are failing. We get so many job applications from people...
When they were at [$]30 [per share] I don't know exactly what [Sun Chairman and CEO] Scott [McNealy] offered but it was somewhere in the mid-20s somewhere. And they would jump at that today -- I mean that's double their current stock price.
And they apparently have a very pretty, new PowerBook coming out that they will finally price at a reasonable price. But their marketing people, all of their engineering people have just been in a cloud ... They have no grip on reality.
About the author
Rob Guth is a correspondent with the Tokyo bureau of the IDG News Service, a SunWorld affiliate. Reach Rob at Rob.Guth@sunworld.com.
If you have technical problems with this magazine, contact firstname.lastname@example.org