1 of 2
PAUL SCHINDLER: How will Marimba's work affect the average Internet user?
KIM POLESE: When I said that my most optimistic view of the Internet was that it became invisible, and you just accessed services, like you access electricity, that's what we enable. By that, I mean, what Marimba does, we allow applications to be delivered invisibly, transparently, across the Internet, and updated and managed in the background. An application is really a service like package tracking, or home banking, or travel reservation systems.

The impact Marimba will make on the Internet will be to enable the Internet to be used in a whole new way. Instead of just for browsing, the Internet will be used for accessing services that are comprehensive, compelling, easy-to-use -- that don't require any technical knowledge or downloading -- that are personalized, customized -- that basically deliver on the whole Nirvana version of the Internet that we've all heard about. That's what our technology will enable, used through our customers, by developing applications and delivering them.

*
"Marimba will ... basically deliver on the whole Nirvana version of the Internet that we've all heard about."
*
SCHINDLER: Who's right in the Microsoft vs. Sun suit over Java?
POLESE: I can't claim to be an expert on this debate because I have not read the contract. However, I do very strongly believe in the need for a platform independent implementation of the Java Virtual Machine, and compatibility between Java Virtual Machines. Whatever it takes to make that happen -- and I think it means the entire industry rallying around and exerting pressure upon everyone who is driving the future of Java -- I am in support of.

I do think we as an industry need to make the Java Virtual Machine, specifically the class libraries surrounding the Java Virtual Machine, as simple, compact, and cross-platform as possible. In other words, have a very basic set of class libraries that comprise the fundamental core class libraries, and then libraries that take advantage of a platform can be bundled along with an application. That way you get the best of both worlds. You get a very portable Java Virtual Machine that runs on all platforms, and you can take advantage of the platform you are on by bundling the libraries with your specific application. C O N T I N U E D . . . 2 of 2
SCHINDLER: What are the consequences of failure to achieve portability?
POLESE: Well, because the Internet is driving the proliferation of multiple platforms -- all sorts of different devices will plug into the Internet, including cell phones, PDAs, laptops, all sorts of different servers -- it is essential that we have standards that can operate in a very heterogeneous world. The world is only going to get more heterogeneous, not less heterogeneous, especially with everything having an IP connection.

*
"The reason why PCs have only penetrated half the U.S. population is ... that they were incredibly difficult to use. ... You needed to be ready to install software and configure your DIP switches."
*
The consequences of not having technologies and standards that support cross-platform applications are very bad. It would be a very bad outcome for users. We'd be in a situation where things didn't work everywhere. We'd have constant incompatibilities and clashing.

The end result would be that the Internet would not be a consumer platform. It would not be something the average person would plug into and use. Consumers are not going to put up with having to make a choice between platforms, buying different software for different platforms.

The reason why PCs have only penetrated somewhere around 50 percent of the U.S. population is, in large part -- yes they were expensive -- is that they were incredibly difficult to use. Very complex. You needed to be ready to install software and configure your DIP switches.

If we don't overcome the platform issue with the Internet, then the Internet will never be as successful as I think we all want it to be.

SCHINDLER: Are you optimistic the Industry will get it together and support portability? It didn't work for Cobol or Unix. Do you have any reason for hope in this case?
POLESE: Yes, I do. For one thing, I think we have learned lessons from the experiences of Cobol, Unix, and other technologies. We know what to do and what not to do.
*
"You didn't have Unix running in everybody's cell phone."
*
Also, the economic promise the Internet holds out for all hardware and software vendors is tremendous. People don't want to screw this up. They want to make sure what they're building will work on a whole bunch of different platforms. They don't want to narrow their market, the penetration into the ultimate user base of the entire world's population.

A lot more pressure will be exerted to insure that technologies run in a cross-platform environment than was ever the case before. You didn't have the scenario of Unix being something that ran in everybody's cell phone. You do now have the potential for the pervasiveness of the technologies to every person, literally, in the population.

That increases the pressure and the need for portability.