Ken Arnold's Blog

Community: Jini Archives

Knowing What's Where

So far today the most interesting talk I've been to is the one on RFID techonology by Sun. They've built it on Jini and Rio, which itself is built on Jini, and they have an impresive platform for dong stuff with RFID applications. The match is good because these systems require customization and smarts at the edge of the network, where the data is gathered, because otherwise too much data would be sent back to the central processing systems. Jini is built on the idea of moving behavior to where it's needed. They also have to deal with failures all the time, so having something that is low maintenance to install and heal is critical.

What's going on with RFID itself is also interesting. We're getting to the point where tracking objects -- real objects -- will be a major focus of computing. When you can know where every piece of paper is in your files we will have a new scale of relationship modeling. (Well, new to most -- in the past anyone dealing with this much specific information was using large-scale systems that were very hard to build.) I think there will be a lot of interesting work in figuring out how to correlate all the information in a timely fashion.

Tomorrow's candidate for "Most Likely to be Wow" is the Orbitz talk. Ever thing you book on Orbitz runs on a Jini backplane.

The Unexpected Newborn Adult

JavaOne is next week. Of course you know that.

The folks who do Java stuff out in the world have ranged from intrigued to fanatic about Jini and JavaSpaces, but those who set the schedule always seem to have something else to talk about. Somehow the Jini/JavaSpaces stuff is always "new". It's been "new" for over six years now. Last year it was part of the emerging technology track over five years after it emerged. This year it's apparently "unexpected". I kid you not: The track it's in is "Intruiging and Unexpected: New and Cool".

Let's get real. Jini is in mission critical web sites (this year's talk by Orbitz might get your attention, for example). Companies have been run for years on its software. Sun itself is betting its RFID strategy on Jini. Large scale systems use JavaSpaces, starting with its introduction six years ago as a show-wide demo, and including (this year) a a large grid-computing system by a major investment company to follow on from last year's talk by Freddie Mac, among others.

The one talk you must go to if you have any doubts is the overview of where the whole technology stands. Your head will spin as you are whirlwinded through a survey of what's actually happening in the marketplace, both in industry leaders and startups. Then make up your own mind.

But to Sun this stuff seems to be perpetually arriving. New. Every year it's new. Which is unexpected. You'd think Sun would have noticed by now that it isn't new, that Jini and JavaSpaces are being used in ways where, if they didn't do the job, they would be chucked out before the company failed. Intruiging, isn't it?

For many folks at Sun there are two kinds of Java: J2EE and J2ME. Everything else is just baggage. And these folks seem to control the agenda for JavaOne. The number of rejected talks by big projects using Jini, JavaSpaces, and other Java technologies is pretty amazing.

The Jini community -- inside Sun and out -- keeps sticking to it, thank goodness. This year, as in the past, we'll just make our own conference-within-a-conference. Your primary resource for this is probably Frank Sommer's summary of what's shaking with Jini and JavaSpaces at JavaOne this year. There are a lot of things going on, and more things happening informally. Keep your eye out for the Jini folks, both from Sun and otherwise. There will be many around.

Maybe it's time we started passing out "Ask Me About Jini" buttons. Start our own pyramid scheme. If Sun won't bother to make Jini and JavaSpaces visible -- even squelch them in this most public Java forum -- we'll just have to make it happen ourselves. Join in, or at least give the overview talk a listen to see whether you ought to care. I think you'll be surprised how this perpetual newborn is already a responsible (if cool) adult.

JiniFest

Last night was JiniFest, which was a blast. JiniFest is what the Jini community does at JavaOne to have folks show what they're doing with Jini. Some are community projects led by interested geeks and others are commercial products. So while the Java community folks are doing boring (if important) process stuff, the Jini folks are having a party.

Jini is generally the best kept secret of Java, and in that vein, The Powers That Be™ put JiniFest in the farthest possible room from the main area of the community event. You had to walk past empty rooms to find it. Not that this implies that some folks in Sun are putting Jini down, of course.

The room, though, was crowded with folks doing fun stuff with Jini. If nothing else, we had the largest percentage of Brazilian of any event on the floor. There were around 30 community projects and companies with tables showing what they've been doing. Some have been with us since Jini was released, and others just went public. At least one company went from no Jini code to full product release in less than nine months, so it all happened since the last JiniFest.

I can't possibly remember it all, and I don't want to play favorites. Some companies had Jini development and deployment tools, including some interesting hardware demos with stuff that looks like props from an Austin Powers film (that is, it looks like a spoof on 007 technology). Several companies had products using Jini to build systems, including financial and taxes to security. Some were so new they hadn't even started last year, and others are ongoing for some time. The projects were led by some folks who just had something they wanted to do.

The first annual award for service to the community went to Bill Venners, who (among other things) runs artima.com. Bill has been doing the monthly FAQ postings, drove the first approved community standard, has been on the oversight committee since the beginning, etc. In other words: Well, duh, who else? Congrats, Bill.

And the message is...

How many guest speakers from how many big companies do we get? A lot of what's been exciting in the past were talks from companies we've never heard of doing things we've never thought of. This year, at least today, we've been treated to speakers (yes, sometimes fun) only from Sun's major corporate partners. Can we at least spread this out a bit?

Yet the message is fun. It's great that Sun starts bringing together the notion of a single platform. There are incompatibilities between the platforms that need to be smoothed, including some that never needed to be there. It will be nice to pull them together. These incompatibilities have been a thumbtack on the floor for network things like Jini all along: What can downloaded code rely upon?

Still, I confess to be baffled by why we need a single network. One of the things that Jini has proved is that you don't need that: You need one platform so I can give you code that knows how to talk to me, and then the network becomes an implementation detail.

Already today, there are people using Jini to build networks that simultaneously and seamlessly run across Bluetooth, ethernet, LonWorks, and X10. Why one network?

In effect, Jini is the object layer on the network stack. We agree in advance on a Java API, instead of agreeing in advance on network protocols. The implementation gets filled in as appropriate for the local network environment.

"Everyone, Everything, Everywhere, Everytime." So why "One Network"? How about using that One Platform to build on "Every Net"?

This work is licensed under a Creative Commons License.