One more cup of coffee for the road.
I've never attended a JavaOne conference where the final keynote has failed to pique my interest and this year is no exception. James Gosling always comes up with some interesting stuff. This year he featured the GUI builder (Matisse) that looks very slick; a mobile app development tool capable of debugging an app while it's running in the actual phone; and a small (40 pound) unmanned aircraft built by the Boeing Corporation whose pilot is a Java program.
John Gage showed a demo of the ultimate in code mobility. Processes that can be moved from one machine to another in mid-stream execution. He showed two programs - one that prints an ascii image of Duke to standard out and the other that does the same for the Java Logo. He started each process on a different computer and then moved one process in the middle of execution so that they both were running on the same computer.
Of course this raised some questions in my mind. What if the processes are writing to a file instead of standard out? Is a new file created on the target machine or is the old file still accessed across the network? If the latter, what happens if the file is not network accessible from the other machine?? Just details I suppose, important details, but details none-the-less.
The keynote ended with a panel discussion moderated by John Gage and consisting of James Gosling, Danny Hillis (CTO of Applied Minds), Bill Joy, Paul Saffo (Director of Institute for the Future), and Guy Steele. Appropriately enough for the last day at JavaOne, the discussions were mostly about the future.
I epecially enjoyed hearing about a new language currently under development called Fortress. Guy Steele stated that Fortress attempts to do for Fortran what Java does for C++. There were not a lot of details given, but from what I could gather the language will offer some special syntax that faciltates the expression of mathematical formulae.
On the other end of the spectrum, Paul Saffo talked about his greatest fear. What is he scared about? Why Google of course. He states, now that Google has digitized the entire Internet they are beginning to look with a hungry eye at the physical world (satellite photos on Google Maps are only the beginning). He conjures up a scenario in which he it sitting at his desk at home. He hears the pet door open and shut, but it's not his dog, no, it's a GoogleBot come to digitize his home and everything in it. Yeah, it's out there, but it may be that the most successful companies of the future will be those who are able to discern exactly what types of physical world data are most important, successfully digitize that data, and make it available over the Internet.
Speaking of the future, I greatly enjoyed Bruce Boyes talk on "Self Configuring Smart Robots Through Plug and Play Boards". Bruce is the Technical Director at Systronix, Inc.. They manufacture a line of 32-bit real time controllers that execute Java byte codes natively. Just perfect for robot construction (among other things).
Bruce talked about applying a technique in which on board XML (XML tagging memory) is used to configure each robot in a swarm. Configuring each robot uniquely in a swarm of say 2000 robots would require multiple code bases. More code, as we all know, means more chances for something to go wrong and more potential maintenance overhead. By using XML tagging memory, one code base can be applied and each robot becomes self configuring. It's a perfect example of the old adage, "One man's hardware is another man's software (and vice versa)".
Robot swarms are potentially very useful for many tasks. Imagine sending 2000 robots to Mars instead of just two. Survivability goes way up and the swarm can perform more efficient types of analysis through intelligent division of labor.
By the way, Bruce is one of the Community Leaders of the Robotics Community on java.net. So if you're a robotics fan, please check it out.
Well as always JavaOne has been a hoot but my cup is just about empty now. The technical sessions are done, the pavillon is closed down and all the speakers and attendees have gone home for a well deserved rest. The technology and the buzz, however, go on.
Next year we will all have forgotten how exhausting the whole thing has been and we will return to once again to marvel at Java's progress over the ensuing year.