We recently caught up with James Gosling, the creator of Java about his new role at Sun, software patents, the open source movement, and the future of Java.
Ahead of the upcoming "James Gosling Unplugged" Sun Developer Days in Sydney and Melbourne, Brendon Chase talks to the creator of Java about his new role at Sun, software patents, the open source movement, and the future of Java.
Builder AU: What are you doing day-to-day at Sun Microsystems at the moment?
James Gosling: I'm CTO of the Java developer products group, which means I spend way too much time in meetings. When I'm lucky, I get to go and participate in engineering discussions. This is mostly around Netbeans and real-time programming. I also do a lot of developer conferences, interviews, and so forth.
Do you miss the time doing research?
Yeah, but I sort of spent a couple of years in Sun labs just recently and the stuff I was working on got to the point that it was ready to actually turn into a product so I figured I had to go and actually work in a product group for a while for it to become a real product.
Did you ever think back in the days of the Green Project that created Java that it would evolve into what it is today? Did you envision it would be used the way it is today?
Well, it's one of those yes and no things. We sort of sat around and hypothesised all kinds of stuff and actually wrote up a document that was a bunch of use case scenarios that actually got to be some stuff that's going on today and that is why Java has held up so well. But at the time those things were an exercise in science fictionÃ¢â¬"it's one thing to speculate on what things might do but another to see it actually happen. I never actually believed any of the stuff we were doing was anything but science fiction.
Kodak paid a handsome US$92 million a couple of months ago for alleged patent infringements in Java. Being so close to the creation of Java, how convinced were you of their claims? This is one of those things where I shouldn't say very much other than the whole case was a complete atrocity. One of these days we need to write up a pretty thorough account of what happened during the case. But it really opened up my eyes about patents and the American legal system. It was quite bizarre and I've already said more than I should.
Do you believe in software patents personally?
At some level yes, but the way they are being used today is ridiculous.
Do you ever look back on the early days of Emacs and when you sold the rights to Unipress and their dealings with Richard Stallman and think you created a monster?
I don't know about creating a monster. I mean Stallman is something of a monster, I don't know about created. He certainly freaked out when I gave my rights to Unipress. I had gone all over the Net trying to find people willing to take over the maintaining of it, including him, and you know, everybody has a day job.
The reason I was looking for somebody to take it over for me because I had reached this sort of crisis where I had to be either Mr Emacs or I could graduate. I couldn't do both. I shopped it around and nobody said they would take it. Stallman in particular didn't want to have anything to do with anything that looked like UNIX because he thought UNIX was truly an abomination, but Unipress, who were two guys and a garage, were willing to take it over and they charged a very nominal feeÃ¢â¬"enough to pay for lunch and the rent. Then they got screwed by Stallman and his grand principles and I think that is deplorable.
Is the animosity still there between you two?
Oh, I think he is a nutcase.
What are your thoughts on aspect orientated programming and a language like AspectJ?
I have a love-hate thing with Aspect J. There are many things I like about it. In particular it is very easy to destroy the invariance that programmers expect when they define an API because it allows a third party to wiggle their way between this relationships between two sides. The way that it allows people to violate that contract I find actually fairly disturbing. It's got that butter knife problem, you can use it either for good or evil and I have to wonder how often it would be used either way. It's the kind of tool I'd like to use but would hate for somebody to use at me.
Do you ever worry that the open source world, or IBM in particular, will make an implementation for the Java platform and overshadow the current offering from Sun?
I find it funny you say open source world or in particular IBM. We are very involved in the open source world. I'm an ex-IBM employee, so speaking from my memory they are the least open source company I could imagine. It's certainly possible to imagine somebody doing that but we are very open with our dealings with the community. There are some people who are grumpy about it but we try and balance the desires of the community. On the one hand they want more openness, and we try to be incredibly open, and on the other hand they want consistency, reliability, interoperability, and a lot of things that conflict with full open source licenses. In most ways we are completely open source, all of our sources are up on the Net, you can download it anytime you wantÃ¢â¬"it's really easy to find.
What's the next big thing for Java?
Well, one of my standard answers for that is -so what do you want to be next"? Java is run through the Java Community Process so if you want to see something in the next release, get involved and go to www.jcp.org. There is a way to enter the debate by making suggestions or voting on suggestions and being involved and guiding it. Java is a community.
What has the feedback to Java Studio Creator been so far?
Pretty overwhelmingly positiveÃ¢â¬"people are loving it. It's enabled people to rapidly create these Web applications that on one hand you can put together very quickly and on the other hand are able to live in between the full J2EE framework which gives them massive scalability, reliability, and all the other things that come with J2EE, like giant clusters and so on, which [are] pretty hard to talk to.
What do you think will be the three important IT breakthroughs in the next 10 years?
There are some big ones that feel like they will happen in the next decade. The full broadening of the network to all kinds of devices on the fringe such as automobiles. Automobiles on the network is happening very quickly.
Then there is the final solution to the -last mile" problem that right now a bunch of things people are trying to do has been hampered by broadband not being everywhere. A decade from now it will be everywhere, and then a whole lot of things will then change.
That brings me to number three which is VOIP. VOIP will take over. When I say VOIP, I mean two versions of VOIP, video over IP and voice over IP. We are pretty close to where video IP works pretty well at home. If you have a decent DSL connection you can do pretty nice video.
That is going to be very disruptive to phone companies and to the television companies. The television companies already don't know how to deal with the network. When you've got these digital video recorders like TiVo and Replay, that kind of capability will dominate. The whole notion of commercial television as we know it with commercials and all that will change pretty dramatically. It is already under fire.
Do you think a lawyer will become part of the software development team of the future?
[The lawyer] already is. That one is past tense.
At the moment you are using an Apple for day-to-day use. Can you see a day when you will be using Solaris on x86?
One of the problems with Solaris on x86 has been that support for laptops has been very thin. With Solaris 10 there is a lot of laptop support. I'm currently looking to get myself an x86 laptop but I want to get the right one because Apple notebooks are just physically better, they are mechanically more solid than most of the PC ones. I put too many files on my laptop to deal with something that is going to break and every PC laptop I've ever had is a piece of crap, technically speaking.