How it all started
What inspired you to start Marimba?
I had always wanted to start my own company. It had been this
life-long dream. I always knew that the right opportunity, the right set of
people, and the right product would present itself at some point. At the end
of last year, it became clear that it had happened and I would be an idiot for not
for going and doing this now.
I had been product manager of Java at Sun Microsystems and I had been
working on the product for three years. We succeeded in our goals of getting
it out to the market and making it wildly popular. I had the opportunity to
go off and start this new company now with three of the best engineers on
the Java team. So that was a big factor. The right people.
Then it was the right time in the market. The platform had been established
and what was missing were all the tools and applications that needed to be
built on top of that. Also, the financial landscape of the marketplace for
starting new companies right now, especially Java-related companies is
really good. So it just became obvious that there was this confluence of
events, timing that was happening, and now was the time to go. And we know
kind of what needs to happen next.
So you're saying you were lucky?
Yeah, and we had the right timing. I think it's kind of the ultimate
success of Java to see people splinter off from Sun and start new companies
because that's really a validation of the fact that Java has become a
How did Sun feel about you taking off with obviously some of the
finest engineers they had?
Well, I think Sun, of course, was a little surprised to have three
of us leaving at the same time. The fourth cofounder came from Starwave but he had been at Sun also
previously for a long time. So it was a surprise. It was a shock, but they
recovered quickly because I think they really understood that this was in
fact a good thing for Java. The more new companies out there springing up,
building products on top of the basic platform, the more chance that Java
will truly become ubiquitous. Scott McNealy, the CEO of Sun, was the first
person that we went to and told we were leaving. He was instantly
supportive. So he gets it clearly. He's been an entrepreneur himself.
So you said you know what needs to happen next. What does need to
Well, we had some ideas for products that we wanted to build on top
of Java and we're focusing in on one of those and building it.
[Java] is like real estate in California in the 1950s or something
right now. There are so many opportunities to focus in on one particular
tool or application and build a killer app.
I know you can't talk about that specifically right now, but can
you give us a general idea?
We're building tools for Java programmers. The area that we're
focusing on right now is moving past what you can accomplish with simple
applets in browsers today and enabling more full-blown applications, not
just applets, but applications so that you can do more sophisticated things
than you presently can.
When do you expect to have the first product ready?
We're shipping in October. We're putting it out in public beta on
the Web site; it's actually out in alpha right now. It's been out since May
to about 25 companies who are using it and building applications with it.
Let's talk a little bit about Kleiner Perkins Caufield & Byers and your
relationship with them. Were you instrumental in helping their Java Fund get
The Java Fund? I had heard about it months ago and heard little
whispers about it. We chose Kleiner Perkins as our venture partner because
we felt that their personality and philosophies matched best what we needed
in a venture firm. For instance, strategic alliances: they're very strong in
that area. Their focus on Java was a big plus for us. We didn't have
anything to do with starting the Java Fund per se, but when we found out
about it and Kleiner's interest in us, I think there was a synergy there
that made just a lot of sense.
Kleiner is one of the more well-known venture firms of the Valley, and
seemed to be the obvious choice, but we talked to a lot of really good
top-tier venture firms before making that decision. It was tough because
there are a lot of great firms out there. It was very difficult to narrow it
down to one.
You don't often hear people having the luxury of choosing their
I know! It's kind of an unusual situation, and during the whole
process we kept looking at each other and saying, "It's not supposed to be
like this. It's not supposed to be this easy." But again: unique time,
unique situation. So we realized this is not normal.
What do you think about the whole controversy of the hostile
The thing that's really ironic about it, that I find really amusing
is, that people download binaries, DLLs, plug-ins, and what have you every
day of their lives and never think about it. But that code (that's C
or C++ code or whatever) is so open to hacking, so much more than Java! Yet
people never think about it. Java's a secure programming language. It's
much, much more difficult to introduce a virus into a Java program than it
is into a C or C++ or Visual Basic program or what have you. Java was
designed for building applications that would fly across networks and that
could be open to being hacked by random bad people out there.
So things are missing in the language that makes it really easy to introduce
viruses. Like, there are no pointers in Java. The scares in the press have
come because there are some researchers at Princeton University who have
been sort of making it their life's goal to uncover as many potential
theoretical holes in the security model of Java as possible. They've been
doing a great job. This was in fact the intent of putting Java in source
code form out on the Net. It was to get people working with Sun and working
with the Java community in general to iron out whatever needed to be ironed
out. That's what they've been doing.
What's interesting is that no one has to date introduced a virus in a Java
applet or application. No one has done it. No one has hacked into a Java
applet. I think with all the headlines and all the scares in the news about
security in Java, that message gets dropped. That's extremely relevant
because there are viruses introduced every day into random plug-ins or DLLs
that get downloaded across the Net.
So the bottom line is Java is the most secure programming language,
bar none. And if you're going to be sending code across the network and
you're not encrypting it, Java's the way to go.
NEXT: Life, the universe and everything